Skip to main content

On Platform Design, Part IV: Social Platforms as Aesthetic Generators

The fourth of a six-part essay by design theorist Yin Aiwen on communicative labour and productive architectures in the digital space.



Yeyoon Avis Ann - Post Image of R∞pt artwork

Yeyoon Avis Ann, Post Image of R∞pt (2020). Image courtesy the artist.

1. The Flavours of Social Platforms

It was 2015 when I ran into the question on Zhihu (Quora’s counterpart in China): what are the general structure and critical elements in a bowl of Zhihu-flavored “Chicken Soup for the Soul[1]?” How does it differ from the soups of other platforms?[2] A user named Wan-wan (丸丸) gave an insightful analysis of the types of viral chicken-soups circulating on China's four major social platforms (WeChat, Douban, Weibo and Zhihu) and how they are deeply linked to the platforms’ positioning. 

According to her, the WeChat “chicken soup” instigates an almost hysteric yearning for success that enables excessive hard work to the point of exploitation. According to Wan-wan, the popularisation of this kind of “chicken soup” mainly came from the nature of the platform itself. WeChat is predominantly a venue of professional and family networks, which means it reproduces the traditional hierarchy of Chinese culture. The platform also amplifies the voice of a senior generation who still have fond memories of the 1980s and 90s when striving was always rewarded[3]. Naturally, this generation is prone to sharing stories that validate such an ideology. Sharing such stories helps users attain likes in their peer network, but the sharer also presents themselves as an idealised hard worker.

Zhihu is a platform started by successful professionals from different fields, later expanded to a broader social platform centred around questions and insightful answers. Being part of the literal knowledge economy on Zhihu, its users tend to produce content that implies uniqueness is the pathway to success; the platform’s “soups”, to that end, instigate a feeling of self-appreciation within users. 

Yeyoon Avis Ann, Screenshot of a tracklist by Bedok Bandits from R∞pt

Yeyoon Avis Ann, Screenshot of a tracklist by Bedok Bandits from R∞pt (2020). Image courtesy the artist.

Douban is another social platform for sharing books, music and films that usually attracts young, liberal-leaning users. Its “soups” typically provide a taste of nihilistic cosmopolitanism, such as the sense that since all beings on earth fundamentally suffer, humans can only feel pleasure in mere degrees, and success is an illusion.

And finally, there is Weibo. For Wan-wan, the biggest public-oriented social platform in China, which gathers the broadest demographic of Chinese netizens, it tends to produce and circulate cynical “soups” that are almost anti-chicken-soup. The circulation of this kind of “soups” on the platform has a lot to do with the fact that most people aren’t successful yet cannot conceive of success outside society’s expectations. Thus, mocking everything becomes a popular way to neutralise their collective disenchantment perhaps.

According to art critic and media theorist Boris Groys, “Today, everyone is subjected to an aesthetic evaluation — everyone is required to take aesthetic responsibility for his or her appearance in the world, for his or her self-design”[4]. He observed this phenomenon in 2009 when the post-Facebook era had already embroiled everyone into the global theatre of social media. Therein everyone is subjected to the gaze of others, and one must stay vigilant about maintaining their self-design to manage their image. In the following year, Instagram came to life and swept over the world. It became even more accessible for people to design themselves through abstract photos, atmospheric filters and discrete hashtags. One no longer needed to process their thoughts and opinions, nor did they need to engage with a community context. A selfie or a food pic would be a good enough gesture to project their ideal self-image. 

Social platforms — the architecture that centres around the self and exposes the self to a curated context — are our present-day theatres.

Social psychologist Erving Goffman used the term “dramaturgy”[5] to describe how, in social situations, individuals presented themselves as actors performing in the theatre of society. People do so because they want to control how they are perceived, manage their image, and the kind of attention they want to attract. Each gesture, interaction, choice of words and detail of a person’s image is curated, consciously or subconsciously designed to fulfil their idealised image of themselves. As with many aspects of the human experience, such performativity and self-design were amplified through the Internet, although Goffman proposed his theories before the Internet existed. 

But such performativity cannot be staged without the existence of the “theatre”. Constructing this theatre, we find both one’s internal framing of their idealised self and the social context that conditions the interior structure. Social platforms — the architecture that centres around the self and exposes the self to a curated context — are our present-day theatres. From their initial positioning and marketing strategies, these platforms lay out the nature of the network they are constructing. Even though an individual may contain a multitude of personalities, in the course of their search for the self and their different ways and modes of engaging with others[6], one’s presentation of the self is always responsive to — if not conditioned by — the social environment, network or the platform itself. Children will filter what they talk about in front of their parents; an employee will think twice before posting something visible to their boss; friends consider social influence before sharing content. Perception management is deeply attached to the networks where the user is currently present. The content that people produce aligns itself with the disposition of the platform. At this point, the model of cultural production in cyberspace has moved far from its initial subject-centred production, proliferated with ideologies and aesthetic varieties, to now quite simply being a digitalisation of Erving Goffman’s theory of self-presentation. 

People now relate a platform to a particular aesthetic rather than the individual presence or expression.

If how we understand ourselves and our relationships with others is crucial to our communication and creativity, it suffices to say that the source of the communicative products has already been trimmed and disciplined the moment we get on a social platform. Platforms put everyone on designed stages, guiding users through a tightly-curated process wherein people no longer need to explore their alternative selves and keep their relationships close. Moreover, platforms accelerate people’s gratification for immediate attention by offering templates, filters, stickers and algorithmically-adapted tools which ensure an efficient and universally appealing outcome[7]. With that, when users subject themselves to the gaze of others, platforms can arrange a perfectly closed loop of communicative production and automate most of the process. Social platforms have thus become de-facto aesthetic generators, through which users — and their content — are presented as perfectly desirable to the network they have constructed. 

Consequently, the conditions these aesthetic generators create lead to an aestheticised collective subconscious. Each platform has a set of expectations, which compel people to produce content and interact so that the particular user network would appreciate it. People now relate a platform to a particular aesthetic rather than the individual presence or expression. Tumblr is about the post-Internet, Instagram is optimistic luxury with nostalgic filters, and Twitter continues to be “obnoxious and negative”. 

Even if platforms are not interested in differentiating themselves from each other interfacially, they want to distinguish themselves in features and user temperament — a delicate balancing act. It is the old chicken-or-egg question: is it the platform that produces the aesthetic, or do users who make a certain aesthetic tend to be drawn to a particular platform? Joining a platform today is much like going to a new town; not only do you need to familiarise yourself with the interfacial environment, you also must adapt to a particular culture to communicate, exchange, and so on. The challenges of UI/UX (User Interaction/User Experience) have resulted in a decline of new platforms and initiatives and a tendency for surviving platforms to polarise themselves so that they are distinct from each other in terms of the content they circulate and their community vibes[8]

The question is, why and how did platforms become such “branded stages” in the first place?

Twitter screenshot

"Twitter is about being obnoxious and negative", a tweet by @DISmagazine, now a suspended account

2. The Market, Platform Positioning and the Making of Aesthetic Generators

While this may be news to many people nowadays, Sina Weibo was not the only Twitter counterpart in China in the late 2000s. Many duplicates were from tech moguls like Tencent and Sohu or small independent teams like Fanfou and Jiwai[9]. Among many competitors, Sina got a breakthrough with its celebrity strategy and successfully expanded the market by attracting a wide range of users who were excited to speak directly with the big names. Many companies followed suit, but all failed. WeChat was the only major social platform that got a market share in the post-Weibo Chinese Internet, not by directly competing against Weibo with a platform that was the “same but better”, but by opening a completely different “theatre” with its nature. If Weibo offered a public venue that aggregated personal blogging and interest communities through hashtags, trending and Super Topics[10], then WeChat offered a predominantly private venue for people who already knew each other. It uses the same strategy as Facebook: reproducing people’s real-life social networks to the online sphere, so their users can always begin their journey on the app with attentional rewards[11].

Yeyoon Avis Ann, Post Image of R∞pt

Yeyoon Avis Ann, Post Image of R∞pt (2020). Image courtesy the artist.

Pony Ma, the founder of Tencent, concluded the success of WeChat after Weibo with this statement: “there is no way to win over your competitors with the same product, the only way (…) is to find a completely different product”. Marketing 101 tells us that every product needs to find its unique position to survive in the market, for the exact reason Pony Ma describes. For platforms, this is an even more critical principle. A platform’s success essentially relies on the richness of the network it can capture: the number of people and entities it accommodates, the sum of information it gets to circulate and the effective connections it manages to facilitate. For this, assessing costs and benefits is critical. Platforms have to make sure most of their services are affordable (free, if possible) so that people stay online. They also need to keep people interested through the unceasing push of content to ensure the algorithm captures people’s attention. Criticality and thoughtfulness are challenging to achieve consistently and undesirable as they slow down circulation. As platforms conflate the positions[12] of consumer and producer, they also need to create ongoing feedback loops to incentivise production and consumption. Features such as algorithmic recommendations, constant notifications, and auto-correct, auto-complete or auto-beautification significantly accelerate production speed and are thus desirable to have. 

These aspects ensure that platforms are costly to build from the very beginning and that — naturally — every platform would want to become a monopoly of its kind. For many other businesses, being a monopoly might allow them to be on the throne of the market. Still, monopolisation is a matter of life and death in platforms' business. But there are only many platforms that are lucky enough to capture the right moment and manage to become an all-encompassing platform (so much so that even their haters cannot do without them). In other words, after the horizontal market has been occupied, the only thing left for new projects is to mine the vertical market. 

For platforms, there are three ways of attaining the monopoly position. Some go after celebrity brandings, such as with Justin Timberlake for MySpace[13], Taylor Swift for Apple Music[14] and Jay-Z for Tidal[15]. This trend died fast. Unlike Sina, Weibo only used celebrities as exemplary users, whereas celebrity branding attempted to set up the aesthetic tone. As it turns out, remaining aesthetically neutral seems to be the most appropriate and functional choice for a platform. Practically speaking, people expect to use platforms to share their points of view; no one — including fans of a celebrity — wants their opinion represented by other people. To borrow sociologist Luc Boltanski’s term, the platform is an institution expected to have a “bodiless point of view”[16], so that their users can work on their embodied persona from scratch and be presented fairly. 

Once users tap into an outlet, they are continuously subjected to the platform's gaze crafted by its market positioning. Users are relegated to tiny gears inside a giant aesthetic generator.

Other platforms provide whimsical features for new ways of stimulating or producing modes of interaction. For instance, Snapchat created a sense of exclusivity and liveness[17] among the interlocutors by imposing a time limit for messages, gaining them popularity in a short amount of time. But features are easy to duplicate, especially in the IT industry. The novelty brought Snapchat attention, but attention doesn’t necessarily bring in stable networks. Instead, a plausible network takes time to accumulate and stabilise, but it takes time to dissolve once established. For that reason, connecting to a sense of solidarity amongst users — usually through a particular domain of interest or co-opting a shared sentiment — is a more secure investment. Hence we see post-Facebook platforms like Pinterest, Medium, TikTok and Clubhouse, all of them zooming into interest communities even though variety and innovation in their features differ. 

At first glance, the vertical platforms may appear to have assembled into early subject-centred communities. But no user can negotiate their relationships with platforms like they could with online forums or blogger communities. Not only did people lose the opportunities to define themselves and their relationships with a community, but the only possible engagement between a user and a platform is also in the user deciding if they want to use a platform or not. And given the nature of platforms, dropping out also means that the users have to cut ties with the entire network ecology, which the platform had meant to monopolise. Leaving a platform thus becomes a costly act instead of an option. Therefore, once users tap into an outlet, they are continuously subjected to the platform's gaze crafted by its market positioning. Users are relegated to tiny gears inside a giant aesthetic generator. 

The vicious circle of user-generated aesthetic

The vicious circle of user-generated aesthetic being trained and concentrated by the platform position. Image by Leonardo Landoni, courtesy the author.

The economic logic between subject-driven communities and vertical platforms are fundamentally different too. The development of a platform already requires large amounts of capital that only venture capital firms and private investors can provide; its functioning logic also decides that a platform will not reward context-specific communication as it pursues maximum circulation. The platforms’ fear of losing users makes them intolerant toward unguided experiences and un-facilitated communication. Having complete control of user experience and communicative exchanges permit platforms to maximise their capture of user data that can convert to profit[18], which is the only way to finance the resources that sustain a platform. Despite emerging as unintended outcomes, the repression of community vibes, the deprivation of people’s agency over their relationships, and the new role of platforms as aesthetic generators later became conditions necessary to their survival. From here, we see a detrimental circle closed. 

3. Automating Personas

Behind the scenes is the automation of communicative production, from which the influencer industry, “like” and “troll” farms, and social bots arise. Emotional expression can be reduced to a click of emojis, the state of life can be described by Instagram photos, and social status can be summarised by the number of friends or followers. With the expression and aesthetic limitations that the platforms have imposed, users are reduced to small, framed and generative content bits easily resembled by social bots. This falls into what Bernard Stiegler called grammatisation , which “…reduces the sensory continuum to digital bits, imposing a grammar on their order” (Mckenzie Wark)[19]. As such, cyber personas have become industrialised and easily programmable.

It is beyond ironic that influencers have become a billion-dollar global industry with a complete production line and supply chain, as well as a highly mature business model.[20] Influencer incubators started using Renshe (人设, meaning “character design”) as their principal means of production. Renshe is a model that originates from the Japanese animation and gaming industries, which helped design popular characters. Instead of following the traditional method of crafting a character in scriptwriting, a process that painstakingly constructed the core drives, desires, strength and vulnerabilities of a character, the Japanese animation and gaming industries pooled together popular character features in a database so that designers could quickly assemble a character by drawing random combinations from discrete features. In this way, the industries could directly manage each character’s popularity, merchandise potential and marketing strategies.[21]

Databases have gone beyond a tool for design and decisions — also becoming a punishment for those who desire to explore the multitudes of the self.

Japanese philosopher Hiroki Azuma called this phenomenon “database consumption”, meaning that instead of consuming a grand narrative in traditional storytelling, Otaku people — the digital natives — invested in a database culture in which people only consumed from a database of market-proof keywords that provoked immediate desires. People were no longer invited to analyse and discover characters’ complexities and ponder on the deeper meaning of life. Instead, companies and creators were encouraged to merely dip into the database and customise “new” products that their consumers could quickly ingest. In this sense, the Japanese Otaku culture in the 90s already insinuated a “prosumer” culture in a digital economy, where producers and consumers no longer have a creative distance in which they can surprise and challenge each other. Instead, producers are disciplined by the database and consumers locked into their echo chambers. 

This has also happened in the Chinese influencer and entertainment industries in the past ten years. Both influencers and entertainment stars submit to the Renshe logic: a successful persona should be able to be recalled by a handful of keywords trending on Weibo. Entertainment stars have always been somewhat like industrial products, with fandoms made of hierarchy and community protocols. But never before has the star-making process been so disciplined, fandoms so industrialised and standardised. In the case of the influencer industry, when a potential candidate enrols in an influencer incubator[22], they will be quickly assigned with a Renshe and a business category: what kind of product they can sell, and on what kind of platforms (e.g. content on WeChat, fashion on Taobao, personas on live stream). Not only would their future content be quickly produced by formulas, but their appearances (face, body, dressing style, etc.) would also undergo physical and virtual surgery so that they could always remain “in character”. 

Automated assembly line of personas.

Automated assembly line of personas. Image by Leonardo Landoni, courtesy the author.

The idea of always remaining “in character” has imprisoned entertainment stars by reducing their personalities to a single dimension. These stars are now called “data traffic(流量)”[23]. In the eyes of platform capitalism, they are no longer human beings, only products that attract data traffic contributed by fans. Character design is not just about producing cyber-persona products in quick succession but also about training fans and followers to like the characters, to engage with them and defend them, thus contributing and creating traffic by initiating formulaic interactions. These actions became part of communicative production in the persona economy that generates profits for the celebrity, the incubators, the platforms — but not necessarily for the fans and followers.

We see a very similar logic in more and more industries. Netflix famously produced the hit show Orange Is the New Black using “big data”[24]. The process involved mining a pool of keywords online to find out what their audience wanted to see and then producing a show that precisely matched it. For a short while, both the fan-driven database and big-data-driven production were praised for the entertainment industries' democratisation. But that illusion didn’t last long: instead of creating more hits, subsequent shows on the platform relying on the same production method tanked. Quickly enough, Netflix gave up big data as the primary tool to expand their market and turned to acquiring scripts, talents and overseas markets — the usual approaches.[25] 

As cultural industry became increasingly monolithic and people progressively started to trip over their paper-thin Renshe, users started getting tired of the world of Renshe. They crave more holistic, real personalities on their screens. Yet the harm has been done. When influencers or stars disappoint their audience, followers quickly turn towards the Renshe mindset, accusing their idols of betraying their image and demanding they be cast out. The popularisation of data-based consumption has led to toxic echo chambers where people have become less tolerant of the complexities of humanity, of other people, of the content we see, and of our conversations. Databases have gone beyond a tool for design and decisions — also becoming a punishment for those who desire to explore the multitudes of the self. Netflix might be able to change its marketing strategy, but can our society?

Editor’s note on the works by Yeyoon Avis Ann:

R∞pt is an ongoing project by the artist Yeyoon Avis Ann. In its first phase, R∞pt took the form of a record label releasing visual albums, streaming image-based products like Spotify streams music. As we consider the aesthetic-generating function of social media platforms through Yin’s essay, R∞pt introduces an abstract and poetic approach to viewing images around us. Currently, R∞pt is transitioning towards its next phase as a studio concept label with productions that embody a “lyrical lifestyle”.

  • 1.

    “Chicken Soup for the Soul” is the title of a self-help book series, popularised in the United States in the 90s. The series consists of inspirational stories about ordinary people’s lives, aiming to provide a satisfactory understanding of the meaning of everyday life. As this type of storytelling became viral in China, after being loosely translated and published in Chinese publishers (especially so in the Internet), “Chicken Soup” became an euphemism for comforting, motivating and inspirational texts, which were often framed within a cliched rhetoric and format.

  • 2.

    The original question and answer can be found here: (Link in Chinese)

  • 3.

    The 1980s and 1990s saw the second wave of China’s “reform and opening-up” period, in which the government out-sourced many state-owned industries to private sectors while opening up the Chinese market to international trade, resulting in an economic boom that lasted decades. During this period, while many people fell out of permanent employment and the welfare system, some grappled with the opportunities of the booming market and become self-made millionaires. As such, the social ethos of the time celebrates hard work and great reward, as the society had a lot of space for class mobility.

  • 4.

    Boris Groys, Self-Design and Aesthetic Responsibility, e-flux, 2009

  • 5.

    Erving Goffman, The Presentation of Self in Everyday Life, Doubleday, 1959

  • 6.

    See “On Platform Design, Part 1: From Subject-Centred Communities to Persona-Driven Platforms”.

  • 7.

    see chapter 3

  • 8.

    see chapter 1

  • 9.

    Fanfou was one of the earliest micro-blogging or “weibo” services. It was similar to Twitter, and got shut down due to Chinese censorship, then reopened in November 2010. Jiwai is another notable micro-blogging service.

  • 10.

    Weibo Team, Weibo, “What Are Weibo’s Super Topics?”, April 20 2019,

  • 11.

    see chapter 2

  • 12.

    see chapter 3

  • 13.

    Josh Halliday, The Guardian, “Justin Timberlake Buys His Own Social Network with Myspace Investment”, Jun 30 2011,

  • 14.

    Chris Welch, The Verge, “Taylor Swift’s 1989 is Coming to Apple Music”, June 25 2015,

  • 15.

    Andrew Flanagan, Billboard, “It’s Official: Jay-Z’s Historic Tidal Launches with 16 Artist Stakeholders”, 30 March 2015,

  • 16.

    Boltanski, L., 2011 [2009], On Critique - a Sociology of Emancipation, Cambridge (UK), Polity Press.

  • 17.

    see chapter 1 for the importance of liveness for a community

  • 18.

    Zuboff, S. (2020). The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power. PublicAffairs.

  • 19.

    Wark, M. (2017). TL;DR: This attention economy needs Work. . . Versobooks.Com.

  • 20.

    According to a report from Guanyan Baogao, China’s influncer industry has reached a market scale of one hundred billion yuan in the year of 2020.

  • 21.

    H. Azuma, Otaku: Japan's Database Animals, U of Minnesota Press, 2013

  • 22.

    Among an all-encompassing ecosystem of the influencer industry, the influencer incubators are at the base layer, providing unceasing sources of new talents for the market. They function very much like startup incubators, except that they incubate influencers that may eventually lead a personal brand or e-commerce business. These incubators scout talents, design personas, craft marketing strategies, manage and grow social profiles, bridge resources, and so on.

  • 23.

    Qian Zhang and Keith Negus, 2020, “East Asian Pop Music Idol Production and the Emergency of Data Fandom in China”, International Journal of Critical Studies:

  • 24.

    Jessica Twentyman, Diginomica, “Forget Orange; for Netflix, Big Data is the New Black”, July 10 2014,

  • 25.

    Enrique Dans, Forbes, “Netflix: Big Data and Playing A Long Game is Proving a Winning Strategy”, January 15 2020,

Artists and Contributors

 Yin-Aiwen portrait picture

Yin Aiwen

Yin Aiwen is a practicing designer, theorist and strategist, who uses writing, speculative design and time-based art to examine the social impact of planetary communication technologies. She advocates relationship-focused design as a strategy to redesign, re-engineer and reimagine the relationship between technology and society. Besides publishing and exhibiting internationally, she also works as a strategist and researcher for cultural institutions. Yin obtained an MFA degree from Design department of Sandberg Instituut Amsterdam, and a BFA on Visual Communication from Beijing Institute of Fashion Institute. She is a Curatorial Graduation tutor at the Master Institute of Visual Cultures (AKV | St.Joost) and a researcher in Avans University in the Netherland.