From private messaging to personal media streams: Understanding valued features in social media’s third phase

Abstract

Social media has changed tremendously in its quarter-century existence, yet little is known about individual-level use of these services today or the aspects users find most valuable. Drawing from 37 in-depth interviews with an indicative sample of Australians aged 18+, this article adds revealing depth to the survey research and study of young adults that has been the foundation of understandings of social media use. The study’s qualitative, inductive method revealed varied and deliberate use of what are now multifaceted services. Its key findings contradict some common perceptions of social media use: participants value ‘Following’ features that enable them to curate a ‘personal media stream’ tailored to their interests and overwhelmingly choose private messaging tools for personal communication and post sharing over public posting. The article argues that the range of use identified suggests a need for conceptualization of ‘social media’ use that better accounts for the variety of features and also addresses the common use of different services to different ends by a single user. Key concerns of the field, such as the nature of platform power, operate inconsistently across social media service features in ways that are not addressed by singular categorization of social media.

Despite social media’s omnipresence in many lives, we continue to know little about what people do when using social media services and what they value among the features they use. By features, we mean the different experiences available within social media services, such as to observe a feed of followed accounts, explore algorithmically recommended content, search, or message and share content with other users. The expanded range of services and evolution in what users can do with those services means that spending an hour ‘on social media’ can involve an array of personal communication or content consumption. Consequently, we should be skeptical that there is any ‘typical’ behaviour characteristic of ‘using social media’. The range of ways to use social media also has implications for theories about its societal role and for how algorithms and other aspects of ‘platform power’ (Van Dijck and Poell, 2013) can affect that role.

This conceptual article uses evidence from in-depth interviews to illustrate how ‘social media’ has become an overly broad referent that requires division or subcategorization to build understanding of the range of cultural practices it encompasses. The findings present empirical evidence that supports a growing recognition of the need to not frame social media as a monolithic category and recent assertions suggesting subcategories of social media use (Malhotra, 2026; Törnberg and Rogers, 2026). The inquiry behind the research takes a critical media studies approach that is rooted in cultural studies’ inquiry of audiences (Gray, 1992; Morley, 1986) and draws from empirical data about how and why users engage with social media to suggest the need for adjustments in conceptualization of its use. The article relies on established frameworks that allow for users to negotiate their experience of media technologies with the aims of engineers and service providers (as simplified in the ‘circuit of culture’ per Du Gay et al., 1997). It also considers social media as more consistent with many twentieth-century media forms and media use practices than scholarship that has regarded digital media as exceptional. The article reports descriptive findings indicating significant variation in use, features, and experiences valued by users that is beneficial to subsequent research that might use other methods to identify representative patterns of use or more deeply investigate the dynamics of specific features.

The most widely used social media services in the west (e.g. Facebook, Instagram, TikTok) now offer many features that enable access to different compilations of content that can be personally or algorithmically curated, as well as messaging among groups and individuals. ¹ We found many combine use of a multiplicity of services to distinctive ends; any study of an individual’s use bound to a single platform fails to appreciate that range of complementary use. We consequently organize our analysis by features rather than services.

The interview research discussed here found that participants preferred to use Following features (i.e. features showing content from followed accounts, like Facebook’s or Instagram’s homepage) and Sharing features, or messages shared by friends, despite the focus of much scholarly and popular discourse on concerns related to the use of Exploring features (i.e. algorithmic-based features showing content from non-followed accounts, such as Instagram's Reels or TikTok's For You page). ² The interviews also suggest that the ‘sociality’ of social media use is overwhelmingly through private content sharing and messaging. The priority on Following features challenges concerns of hours wasted on ‘brainrot’ while the abundant private messaging has been invisible to research of publicly accessible feeds.

Empirical findings that contradict pervasive assumptions about social media use suggest the need to refine theories that generally ascribe platform impact. Given present variation in social media features and use, theorization must recognize the range of experiences of social media based on the features and services users prioritize; conceptualization will be stronger at the feature level rather than making claims of all social media. Our fieldwork indicates that using social media in the 2020s must be understood to encompass an array of practices and experiences and that it is difficult to claim a ‘dominant’ or common way of using exists.

The evidence from our interviews indicates we must speak of and theorize differentiated uses of social media. Our findings suggest common tropes imagined of use such as frequent posting, trolling, and interest/engagement with news and politics to be far more peripheral than typical. The evidence-based conceptual work that supports the analysis here suffers the limitations of its methods – particularly related to the scalability of the findings and self-report, which have been addressed in subsequent survey research (Lotz and Lunardi, 2026) – but nevertheless evidences and introduces deeper insight into the varied behaviour common to ‘using social media’.

Social media use is now well established and used by many across the population of many countries. This fieldwork investigates Australia, where recent government survey research revealed more Australians used Facebook in the last week than broadcast television (Australian Communications and Media Authority, 2024a; 2024b). The need for in-depth accounts of how social media is used, why users value it, or what they value is evident as governments around the globe, including in Australia, introduce controversial limits on social media use. It may be impossible to understand social media use with the clarity that developed for twentieth-century media that offered users much less choice or control, but it is possible to improve our understanding with findings from methods that offer ‘thick description’ (Geertz, 1973) and individual-level insight. The features of media cultures will vary by national context based on the history of media use, technological dynamics, and economic and regulatory factors, but this investigation of a cross section of Australian users reveals patterns of use likely to have relevance beyond its shores.

Contextualizing changes in social media features and use

‘Social media’ has been an object of study for more than two decades but its business priorities, competitive dynamics, and characteristics tied to users have transformed through time. The term originated when capabilities of services were narrower and more consistent, but it has become imprecise as ‘social media’ has come to encompass services with increasingly different features: Discord versus Instagram, for example. It has also been studied and conceptualized quite separately from other forms of media despite growing similarities to the content-driven aims of twentieth-century media consumption. By the 2020s, the assumed coherence of ‘social media’ as a category became increasingly tenuous as the term was applied to a range of services with quite disparate features and watching and reading commercial media content became a common use.

Social media has evolved through three distinct phases that have governed its use and study. Each phase spans roughly a decade and has generally expanded the features, and consequently, available ways of using social media. The phases organize key industrial or structural adjustments (e.g. launching of new services, business mergers or acquisitions), technological changes (e.g. new features), user and popularity shifts (e.g. fluctuation in the number of people or categories of people using services), and content and cultural characteristics over time (e.g. terms used for features, types of content popular).

Phase one, encompassing the 2000s, was marked by the emergence of ‘social networking sites’ such as MySpace (2003) and Facebook (2004) and the use of these sites by young users and early adopters. The unprecedented public communication created optimism among tech enthusiasts and media scholars but also moral panics about the implications of social networks for youth. Research by boyd (2014) and Ito et al. (2010) rebutted the panic with empirical data showing teens’ social media use had many motivations, among them that it was a new technology that was not understood or used by their parents and allowed engaging with peers without supervision. Many of the ideas about social media use focussing on participation and self-expression discovered in this era continue to dominate perceptions of social media use even though the user base that informed the research and the capabilities of social media have profoundly changed.

The second phase of social media, spanning roughly the 2010s, featured its mass adoption, as it became used by more than half the population of the US and similarly industrialized countries (see Olenski, 2013), the proliferation and specialization of social media services, and the initial professionalization of ‘user-generated content’. The expanded user base warranted much greater investigation of social media use beyond that of young people, although studies comparable to those of Ito et al. (2010) and boyd (2014) exploring a broader population of users were not conducted (excepting Boczkowski, 2021). Tandoc et al. (2019) identified ‘platform swinging’, or the use of multiple social media services for different purposes – Instagram for posting photos, Facebook for engaging in groups – but argued such use was a result of the specificity of services’ affordances that is particular to the second phase. Our interviews indicate that no service is used in a uniform manner in a context of much greater feature variability; different users rely on different services for different purposes, not necessarily following any norm.

A third phase emerges in the 2020s and is distinguished by the expansion of algorithmically curated features on social media (boyd, 2026; Gerbaudo, 2026), the consolidation of multiple features within a single service, the homogenization of features offered across platforms, and the proliferation of commercial and professional content on social media. Widely used services offer a range of similar functions that enable different uses within a single service and common ‘encounters’ across services (Lotz and Lunardi, 2025). As major services come to have comparable features, users can have commensurate experiences across services and quite different experiences of a single service. For example, purely algorithmic ‘Exploring’ features such as the ‘For You’ page on TikTok or Reels on Instagram, contrast with ‘Following’ features such as the homepage on Facebook and ‘Following’ page on X/Twitter that fully or partially deliver posts from user-selected accounts. Most popular social media platforms now allow users a comparable range of encounters: to send private messages or share posts/media with individual friends or groups within the service, scroll through short videos, post ephemeral content to features like Stories, or read ‘fixed’ posts on a feed. By the 2020s, Followed feeds and purely algorithmic Exploring features overwhelmingly featured commercial media content produced by sole proprietors, small production companies, large legacy media producers, and major brands. Both the expansion in features and centrality of commercial communication mark an exceptional change from what ‘social media’ was in its first phase.

Research on uses of social media

The study of social media use has been dominated by a reliance on surveys (Alhabash and Ma, 2017; Scherr and Wang, 2021; Sheldon and Bryant, 2016), studies of young people (Alhabash and Ma, 2017; Bengtsson and Johansson, 2022; boyd, 2014; Ito et al., 2010), platform-specific studies (Bhandari and Bimo, 2022; Scherr and Wang, 2021; Sheldon and Bryant, 2016; e.g. Sujon et al., 2018), and the use of computational data as a proxy for audience research, leaving a gap our cross-population and cross-service interviews address. Research on social media use interested in why people use social media services and the value derived has generated motivations or gratifications of use. With rare exception, this research relies on surveys that can only assess very general purpose and fails to measure priority among multiple uses (e.g. Alhabash and Ma, 2017).

Interview-based research with an expansive population is lacking, and widely cited studies have notable methodological limits. ³ The comparative absence of in-depth qualitative research aimed at understanding what users do when using social media, why they select particular services or features, and what aspects of the now multifaceted features and services they most value has prevented scholarship from appreciating users’ active and strategic negotiation of services’ affordances and features.

Conceptualizing social media use is further challenged by the steady change in social media features and services available that limit the continued relevance of previous studies. Many of the field’s understandings were developed in eras of social media services and features significantly different from those in place today. Ellison and boyd (2013) have highlighted rapid technological evolution as one of the biggest challenges of studying social media. The last methodologically comparable study of social media use (Boczkowski et al., 2018) investigated phase two norms in which users needed to use multiple services to access a range in features. The conditions in place by the 2020s require that we investigate use of and attitudes toward features and build broader understanding of social media from the components of its use.

Method and participants

To be clear, this article is not primarily aimed at reporting fieldwork evidence but making an evidence-supported conceptual argument regarding the need for approaches to the study of social media that recognize the variation in services and how and why they are used. The evidence supporting this argument is drawn from hour-long recorded interviews conducted over Zoom with a cross-section of 40 adult Australians (37 of which used social media) recruited by a research firm that controlled for gender, geographic location (urban/regional), age, ⁴ and amount of social media use (high, medium, and low) in its quota sample. ⁵ The sample skewed somewhat more educated and wealthier than Australian averages; our analysis did not find particular patterns of use correlated with gender, education, or income. Our sample matched the relative use of different services identified in a nationally representative survey (Australian Communications and Media Authority, 2024a).

Participants completed a screening survey as part of the sample selection that offers descriptive context of the sample’s social media use. Facebook led in response to the question ‘what social media service is most important to you?’ Though Facebook dominated the screening survey responses (49%), most explained in the interview that this is because they do so many different things through Facebook, including a significant amount of practical communication through Facebook Messenger or Groups (coordination of carpools, children’s sports teams) and browsing on Marketplace. Few were strongly favourable toward the service; many regarded it as a utility, a similar finding to Sujon et al. (2018) in their longitudinal study of Facebook use.

As an inductive first stage of a large and comprehensive inquiry into Australians’ media use broadly, the interviews aimed to enrich our basic understanding of social media use – what do people ‘do’ when using social media, why do people use different services and features, how do they use them, what features do they prefer, how has their use changed over time – and how social media fits amidst broader media use. The interview first enquired about the users’ ‘media diet’ – that is, their routine of media consumption on a typical day encompassing everything they ‘read, watch, listen to, or play’ – and how it changed over time. We then queried their use of social media, exploring history of use, the array of services used, how participants use services, and what they regard as valuable. Participants then narrated a typical ‘session’ of social media use while screen sharing, what we call a ‘live scroll’ method in which participants demonstrated their routine and then would scroll a feed, contextualizing what we observed and responding to our questions about the typicality of the feed and whether they valued what they were seeing. We also asked about common practices such as posting, liking, commenting, and sharing. Nineteen participants shared about their Facebook feeds, 13 Instagram, 2 Reddit, 2 X/Twitter and 1 TikTok in the live scroll segment of the interview. ⁶ The final stage of the interview posed a variety of scenarios (a 15-min wait; 2 hours to yourself) and asked if they would use media in that situation, what media, and why. The use of in-depth interviews and observation of live scrolling revealed the variation among individual experiences and provided deep insight into more prevalent practices.

The study adopts an inductive approach to the analysis of the interviews, which enabled the researchers to draw conclusions based on patterns and themes that emerged from the empirical data (Tracy, 2020). The authors conducted the interviews and watched recordings of the other’s interviews, developing notes about common themes. The researchers identified patterns in use and then systematically investigated responses by reviewing the transcripts and recordings. Our analysis tracked differences in use routines, such as features used and deliberately not used, preferred, and explanations of why, and we assessed the scale and nature of sentiments across the interviews. Emerging interpretations were collectively discussed and revisited by the research team, allowing for a refinement of analytic insights. Researchers performed basic content analysis of the participants’ feeds shared during the live scroll to appreciate the number of posts, frequency of personal relations, ratio of advertisements/ sponsored posts, and ratio of followed to suggested content in Following features. No computational analysis was used.

Relative to the scholarship and research tradition, our participant sample is robust in size and diversity. Independent of demographic variation, our participants were fascinating in the range of use they presented – all utterly foreign to our own. Self-report methods can suffer from a range of limits and bias. It was clear that several participants were aware of public discourses regarding social media as a poor or wasteful use of time. The loosely structured interviews were conversational, and we used phrasing to diminish social desirability bias. We also phrased questions to help counter recall bias and found the live scroll, which was conducted after an extensive discussion of social media use, to consistently affirm participants’ responses; the live scroll also served as a tool to aid recall. Doing the interviews by Zoom also contributed to setting participants more at ease than would be the case in a public meeting, on campus, or in their homes. The recruiting firm delivered a sample with a range of technological familiarity. Our methods precluded non-internet users, but they would also not be social media users.

Our inquiry was not designed to test any hypotheses, but to build an empirical base that supports iterative inquiry. The value of such interviews is the depth of inquiry. We cannot make claims to the representativeness of findings, but we explored key findings from the interviews in a subsequent nationally representative, recruited survey (n = 2020) that found support at scale (see Lotz and Lunardi, 2026).

Key dimensions of social media use in the mid 2020s

The interviews revealed two dimensions of use that were most pervasive and also negligible in existing scholarship and discourse about social media use: the content experience of a ‘personal media stream’ and the centrality of private communication on these services. Because of the lack of previous qualitative, cross-service, multi-age studies it is impossible to know whether our findings are particular to use in the 2020s or whether the patterns we identify have been in place for some time. Certainly approaches that ‘read’ social media applications, such as the Walkthrough method (Light et al., 2018) identified evolving features and affordances (e.g. Leaver et al., 2020; Zulli and Zulli, 2022), but user research has not been a component of this type of method. We suspect these preferences have long guided social media use but common methods and more narrow research questions – for instance those investigating polarization and misinformation – obscured identification of more foundational understanding about general use.

Personal media streams

The most prominent use among our participants was engaging with personal media streams, a term we use to describe scrolling through their homepages or Following features (e.g. the homepage and Stories on Facebook and Instagram, the ‘Following’ tab on TikTok, the homepage on Reddit), which participants regarded as tailored to their particular interests. Most feeds are driven by interest, for example, sports teams’ fan pages, hobbies, cars, or travel aspirations so that a key use of social media approximates watching a personalized television channel or reading a specialized magazine that offers an individual’s precise mix of interests – often highly idiosyncratic. For instance, Kevin (age 27) talked us through his Instagram homepage feed organized around content related to desserts, football – and specifically his preferred team Arsenal, photography, and Japan. His feed also included relationship-related memes sent by his partner, but most content adhered to the four disparate interests. Donna (45), who works in a forensics lab, follows Facebook groups about post-mortem and embalming processes, as well as pages and groups on Tudor history and craft projects to produce a similarly distinctive feed. Others cultivated a narrower stream. For Ryan, a 22-year-old mechanic, 76% of the followed accounts on his shared Instagram feed featured motor racing, custom cars, and motorbikes. Curtis (60) uses Facebook nearly exclusively to access information about motorhomes and motorhome-based travel.

Cultivating a personal media stream is very different from what was imagined as the key use of social media in its first phase, when social media services were spaces centred around social networking and self-presentation (Baym, 2010; Ellison and boyd, 2013). The use of social media to construct personal media streams makes this dimension of use more similar to twentieth-century media use that was motivated by interest rather than personal communication; this is a common and well regarded use of media, contrary to Törnberg and Rogers (2026) assertion such use should be considered ‘passive’. Personal media streams allow us to see how users exert agency despite algorithms and platform power to negotiate an experience they find valuable. ⁷

Personal media streams enable users to access content about much more specific interests than would have been possible with newspapers or broadcast, or even cable television channels aimed at a mass audience, and the precise topical and tonal approach is crucial to the streams’ value. For example, a curiously large number of English Premier League football fans emerged among our participants. This type of football (soccer) is a third-tier sport in Australia and receives limited coverage in traditional Australian media; the personal media streams enable access to specific content, in this case, detailed information about teams locally and abroad – and often with a particular sensibility: some prefer official accounts, others prefer the irreverence of fan-based coverage – that is desirable to users and otherwise unavailable. This specificity of the personal media stream is central to the current micro media era in which media use is extraordinarily fragmented at a societal level. Despite high percentages of the population being ‘on’ social media and even on the same services, there is little intersection in the worlds that social media delivers. Even in this sample of 37 users, phenomenal variation in Following feeds emerged.

The use of social media Following feeds as streams of personalized media demonstrates how different the current phase of social media use is from its origins. Understanding the experience of a personal media stream as a key attribute compelling social media use allows us to see how users negotiate with the services’ affordances to construct experiences they desire. It also makes clear that a lot of social media use is easily understandable in ways consistent with predigital media use – in other words, as driven by interests.

The personal media streams develop in Following features and differ from Exploring features that are purely algorithmically curated. We found that most participants preferred the Following features that enable at least some personal curation and tended to appreciate the algorithmic ‘suggestions’ that have become part of Following feeds in contrast to the content delivered in Exploring functions. Our survey research (Lotz and Lunardi, 2026) found Following features rated higher than Exploring features in a series of questions that asked about use of different features. ⁸

A few of our interview participants preferred Exploring features to engaging with their Following feeds or identified similar value in both. Betsy (21) reported valuing Instagram’s Explore page because it suggested content that was very specific to her taste that she might not be able to find on other features. She was unusual in our sample in her active drive to use this feature to find ‘new, interesting’ content about interests she defined as ‘niche’, like the ‘Frutiger Aero aesthetic’, a tech design style popular in the 2000s. Rachel (19) and Ethan (26) identified value in how easily they can share relatable content found in Exploring features so that their respective use of TikTok’s For You page and Instagram Reels was strongly motivated by connection with their friends; their followed content was more personally relevant and less sharable.

Private posting and sharing

The second prioritized dimension of use we uncovered indicates that the nature of sociality has changed from what popular discourse and scholarship emphasizing social media’s affordance of public posting lead us to expect. An extraordinary amount of personal communication and content sharing, which effectively curates use for others, happens through social media services’ messaging features.

Most participants (89%) reported using private messaging features (e.g. Instagram’s Direct Message, Facebook Messenger) to connect or maintain relationships with their family and friends – both in one-to-one communication as well as with groups of friends as one of their main, and for many, most valued, uses of social media. For example, Olivia (51) uses Facebook Messenger ‘basically just about every night’ to decompress with her ‘work bestie’, while Tess (28), and many others, talked about using Instagram’s Direct Message for sending pictures and life updates to her family.

The desire to post privately is not something new, but scholarship has not attended to it as a priority use of social media or has regarded it as more typical of texting and messaging services such as WhatsApp (Johns et al., 2024). Early work identified users managing visibility and avoiding potential context collapse (boyd, 2014), but a common perception of social media – drawn from early adopters – has been that posting publicly was a compelling affordance. It is possible that most users have never wanted to post or communicate publicly but were limited by the lack of features that allowed private messaging or limited posts to ‘friend/follower’ connections. We discovered that most sociality happens in what Malhotra (2026) calls ‘bounded social media places’ that are limited in reach, such as private groups and chats on social media. Our finding demonstrates that research based on the analysis of public posts, or that only asks questions about publicly accessible social media features overlooks a dominant use.

Many also used private messaging features for the sharing of media content. Sharing content with friends is important to relationship maintenance that can be understood as quite separate from the gratifications that emerge from accessing content of personal interest. Phoebe (36) explains this type of use is a form of bonding with her friends: ‘the things that I send privately to friends are more things that we can both relate to together as friends’. Similarly, Daisy (42) explained ‘my husband and I – even if we’re in separate rooms, even if we’re sitting next to each other on the couch – just send memes back and forth. And that’s kind of a fun time and an enjoyable time’. Ethan (26) also recounted that he has group chats with friends on TikTok, where there’s ‘one big, long feed of everything that everyone has thrown in, which is always very amusing. […] I’ve got some friends that are avid users, so I’ll pop on and be like, “Oh, look, there is 90 new videos to watch.”’

Most participants did not receive such an abundance of shared content to enable it to be their primary experience. Nevertheless, the frequency with which content is shared and was expressed as highly valued highlights how messaging features are an important and rarely considered aspect of social media use that enables friends to act as personal algorithms. Such sharing has been called ‘pebbling’ in general press articles to connect it with penguins’ act of giving pebbles as affection (Brody and Cullen, 2023; Travers, 2024).

The priority on personal messaging and sharing is also supported in survey research of Australians (We Are Social and Meltwater, 2025). Our survey (Lotz and Lunardi, 2026) found that ‘reading personal messages’ was the most important feature for users (71% rated it as important or very important). ⁹ Public posting still happens but is unusual. Most participants rarely post publicly and describe it as something for special occasions; only about a quarter of interview participants frequently posted publicly and often for professional reasons.

If priority on personal media streams suggests that users negotiate with algorithmic curation to make using social media an experience driven by personal interests, the prevalence of private sharing that allows encountering media content through a lens curated by friends enables a user experience that further counters the algorithmic power of services. Recognizing this transition is important to appreciating the role of social media in culture and how social media is valued by many users.

The use of media for connection purposes and friends serving as human algorithms are not new. Long before social media, friends and family would clip and mail stories from the newspaper or recommend films, albums, and other media. The same can be said about sharing media for emotional closeness or bonding. Digital technologies make sharing much easier, more personally relevant, and the abundance of content makes it valuable to have others surfacing posts that can be the basis of conversation. Although participants didn’t talk about sharing as the reason they opened a service, it was clear that many had become accustomed to connecting with others being part of their social media use. Other forms of media are also used for connection through shared consumption. The biggest difference between well-established practices of consuming media together as a form of bonding and sending videos and memes through private messages is how the latter allows for temporal and spatial flexibility.

Investigating ‘digital native’ use

The personal messaging use among so-called ‘digital natives’, those in the 25-to-35-year-old cohort, warrants specific focus. Members of this group became social media users in their teenage years and are now adults with full-time jobs and, for some, young children. Many assumptions about the future of media use at a society level have extrapolated from their teenage experience, but we found important changes when we asked them if and how their uses of social media have changed over the years. Nine of ten talked about how they would post much more as teens or young adults and expressed a preference for using private messages and features over public posts now. As Nora (32) explains: ‘I pretty much never post to a general feed. I don’t think I’ve posted to a feed in over 10 years. I will post to groups occasionally, but most of the time, I’m kind of just responding to posts […] I’m not just, like, putting random thoughts out into the world’.

Two interconnected reasons emerged for the behavioural shift. The first seems to relate to their transition to adulthood and maturity. William’s (34) testimonial perfectly summarizes the general feeling of a shift of behaviour on social media habits as a reflection of change in life stage: ‘I posted a lot more on social media previously, because I would get a coffee, and I’d post it. I would get a donut and post it. Now I find it silly. Like, who cares? Different age, different stage of life, yeah?’ Many participants in this age cohort shared this feeling of ‘who cares’ about their life updates. Ethan (26) talked about how, as a high schooler, he would publicly post ‘multiple things a day’, whereas now he posts twice a year on average. When asked why, he talked about maturing but also identified that ‘people kind of use [Instagram] in a different way these days’. This finding is also important for indicating the need to recognize the role of life stage in social media when researching young people.

Life stage shifts can’t be disentangled from technological and cultural changes on social media use: the functions available on services have changed and societal attitudes toward the services have also shifted (Burgess, 2015). Nora (32) said that she stopped posting on her feed when Instagram introduced Stories and, now always posts on Stories rather than her feed. Other participants, Kelly (26) and Kevin (27), shared similar change in behaviour. The key difference between posting on Stories instead of the feed is that Stories last only 24 hours and the visibility is often more restricted than a post on one’s feed. A younger participant, Betsy (21), explained that she mostly posts on Instagram Stories because she feels ‘a lot more control’ when using this feature as opposed to posting on her feed because Stories disappear after a while and are restricted to her followers, whereas a post on the feed is visible for ‘just anyone’. For Betsy, that can be ‘nerve wracking, [because] you don’t really know who’s actually viewing what you’re sharing’. Betsy also has two Instagram accounts, one public, where she follows ‘friends and a lot of acquaintances’ and one private, where she follows ‘closer friends’ and accounts about her interests; she only posts on the private account.

This change in posting behaviour was more evident among the 25-35-year-old cohort but not restricted to them. Some older and younger participants shared similar behaviour; Vanessa (72), who used to post photos of her grandchildren when she first started using Facebook, now prefers sending photos ‘through email or text message’ instead.

The priority suggested by our participants on using features that enable construction of a personal media stream and on messages sent or shared privately emerged inductively as unexpected aspects of social media use. Much like how Gray’s (1992) exploration of television use revealed levels of deliberate negotiation of technological features and audience activity, these interviews – including the narration of typical use and textual artefact of participant feeds – indicate deliberate use. Although previous research has identified a variety of motivations for use, the interviews provided significant depth from which to understand how individuals have particular ways of using social media that suggest distinctions important to them that have not been well surfaced by other methods.

Our interviews revealed significant variation in what people ‘do’ – in terms of the features used and prioritized – to an extent that suggests that assumptions of social media use as a common or standard experience are dubious. Just as a ‘healthy diet’ might include a wide variety of foods and ways of eating them, the ways of using and content consumed on social media vary profoundly. The range of content leads to extraordinary variation but examination of use of and attitudes toward features indicates patterns and priorities that structure the use of social media for many.

The observed variation of what users do when ‘using social media’, the prevalence of and priority on personal media streams, and privatization of communication on social media services challenge public discourse about social media use and provide insight on users’ negotiation of the features and affordances of services, even in a phase when social media services prioritized the introduction of algorithmically-curated features. The detailed insight from a diverse sample of users provides significant evidence of deliberate and strategic use that negotiates the strongest forms of curation from the services. Our human-level research delivers empirical grounds for richer understanding of the roles and prevalent uses of social media outside of groups such as teens and early adopters previously studied.

Although algorithms dominate the structures and interfaces of platforms and shape how we use social media services (Gillespie, 2014), many elect to use features that allow more user input into the experience and actively follow other users and accounts to curate their feeds to deliver content about the things that they like and are interested in. This suggests variability in the implications of algorithms based on the features most commonly used.

Conclusion

Many of the expectations of social media use held by policymakers, society, and non-specialist scholars remain rooted in the initial novelty of the services as they emerged in the early 2000s or concerns about new features that emerged in the 2020s. Our interview findings, which are supported by nationally representative survey data, indicate that prioritized use isn’t fully anticipated by analysing apps’ capabilities and affordances. Speaking generally of ‘social media use’ is complicated by the variation across the population of users as well as within an individual’s use. At certain times, a user will investigate a feed with focus, clicking through to linked content or playing a video in a manner consistent with attentively watching a television show or reading a newspaper. At other times, that same user will swipe distractedly through a feed or use a less personalized feature (Reels, Explore) to ‘fill time’ or to distract themselves more than engage. This variability illustrates the limits of surveys that query whether people do different tasks with social media without also enquiring about the frequency or importance of those activities. Terminology about social media features is also inconsistently understood by users, and the conversational context of interviews was effective in ensuring shared understanding of the features being discussed.

Our cross-service inquiry revealed strategic and disparate use of multiple social media services by individuals in a way that suggests the impossibility of speaking of a common or dominant use of social media. It also provides evidence that ‘social media’ is a problematic analytic category for audience/user research because of the range of encounters it encompasses. The use of social media as a personal media stream is more like the use of predigital media such as television than current conceptual frameworks allow. The private sharing and human algorithm use suggest adapted uses enabled by technology and necessitated by contemporary media abundance and fragmentation.

Our interviews suggest the need for more and focused investigation of different social media features (Exploring; Following; Sharing) and for the insight of users to feed back into how we theorize social media operation. Cross-service studies of specific encounters – or comparable features across different services – will be enriched by considering how service affordances shape use within particular encounters. Most of the survey and computational research has not been able to capture the sophistication and intentionality of social media use needed to make sure theories account for how audiences negotiate the technologies and affordances engineers and corporations offer them.

Our findings also revealed significant challenges to succinctly characterizing social media use. Although we came to perceive that the use by those under the age of 25 differed from those older in consistent and patterned ways, we found demographic features such as age, gender, education, or income to provide little patterned insight and great internal variation. ¹⁰ Our interviews that interrogated media use broadly – and revealed considerable insight into the leisure interests and life pressures of participants – suggested that personal traits and broader life and lifestyle considerations are more likely grounds of patterned use outside of the particular behaviour of young adults.

Our data also raises important questions for the field in relation to the gaps in our understanding of social media use that derive from reliance on surveys, studies of young people, and computational methods or digital trace evidence that cannot identify private message sharing that rates as a highly important use. Data counting ‘Followers’, ‘Likes’, and their equivalents do little to help us understand individuals’ experience of social media and the patterns that may exist across users.

The strong indication of varied and strategic use being common may not surprise scholars familiar with reception research of predigital technologies, but an empirical base has been lacking. The findings here are but a start toward more sophisticated conceptualization; much more human-level investigation is needed to substantiate the abundant assumptions about how social media are used and their societal implications.

Our preliminary work suggests we should be cautious in regarding so many different services with different features and affordances commonly; understanding of this facet of media use may be strengthened by subcategorizing that recognizes patterns of use either typical of users, features, or services. Space limits prevented detailed reporting of our analysis of live scroll data, but surprising differences emerged in the composition of Following feeds within the same services (ratio of followed, recommended, and ads) that raise questions that require broader investigation about the variability of experience – regardless of content – within any one service. We also found feeds largely devoid of news or political content to an extent quite contrary to scholarly focus.

These interviews are a first stage of a multi-year agenda that will continue to develop deeper understanding of social media use particularly and in relation to broader diets of media use. Although there is much more to learn about social media use, our agenda also suggests the need to diminish the siloing of social media as wholly separate from and different to other forms of media. Much of the content now consumed bears great similarity to a variety of television forms, talk radio, and newspapers and magazines. Analysing the similarities and differences and incorporating user perspectives regarding what they value offer productive agendas of investigation. Users do not draw many of the distinctions common in scholarly thinking, and understanding media use in the digital era requires conceptualizing use from the vantage of users.