1. Introduction

From the varied and vast critiques of social media, there remains a relative blind spot of critical analysis: the subscription side of social media. The subscription side of social media or paying for social media (YouTube Premium, Reddit Premium, Snapchat+, Twitter Blue etc.) became evident in late 2022 with Elon Musk’s announcement of Twitter Blue. Until then, the subscription services remained a dormant element of social media. Even after Musk’s announcement of Twitter Blue, which caused a media storm (Allen 2022; FitzGerald 2022; Yaffe-Bellany 2022), it is still the case that subscription services remain uncritiqued. From this backdrop, the questions that frame this article are: Why have subscription services remained relatively uncritiqued? Should social media be considered part of a mixed model approach (Fuchs 2020, 136)? And: How does the Social Media Subscription Model (SMSM) provide an outlet for “peak data”? The article has two main sections: first, an attempt to explain why subscription services on social media have remained an uncritiqued blind spot and second, an argument for a new emerging model of social media, SMSM, which, in turn, justifies viewing social media as a mixed model.

This paper’s first section begins with a list of subscription services within social media in order to highlight how common subscription services are. Crucially, although many think that social media subscription services are a new phenomenon, they have been around or debated long before 2022/2023. Establishing the list then probes a question: why have subscription services within social media remained uncritiqued? A reasonable case can be stated that until Musk’s bold announcement of Twitter Blue, subscription services were very much an afterthought and remained outside of contemporary cultural discussion and consciousness unlike other parts of social media. They were reported on and hinted at, but never as frontal as Musk’s attempt. However, a more prevalent point is that within the literature of critical analysis of social media, the consensus of how social media companies make a profit, selling user-generated-data to advertisers, allowed for the blind spot of subscription services to develop and remain unexamined. Individual users paying for social media is systemically overlooked because instead of stressing data extraction, it emphasises the paying user as an important point of economic extraction. This has allowed for subscription services to remain dormant and unexamined and has created a blind spot of criticism of social media. For evidence, this article examines three different examples, Srnicek’s notion of platform capitalism, Christian Fuchs’ notion of the audience commodity (inspired by Dallas Smythe) and Yanis Varoufakis’ concept of cloud capital, to establish the running theme of stress placed on data and how the extraction of data is central to profitable social media. But importantly, this centralisation of data allows for an oversight of subscription services. Unearthing this blind spot is the first case to enhance the claim that social media should be considered part of a mixed model (a mixed model means a company utilising multiple different accumulation tactics) understanding and justifies the emergence of the SMSM.

The second section will develop Shoshana Zuboff’s idea of surveillance capitalism and argue subscription services provide a way out of the tension of “peak data” (Lovink 2019, 7). Zuboff’s notion of surveillance capitalism, provides a useful bedrock for grasping key concepts and mechanisms under which social media operates. Once a useful layout has been staged, cracks and problems within the surveillance model will be addressed: new regulations, scandals such as Cambridge Analytica, Apple’s change in privacy policy, destabilising impacts on democracies, an advertising tax and fines from state actors. However, rarely is the tension of peak data considered as a curtailing self-imposed force of social media. Following the logic of Zuboff’s surveillance model, social media is logically (applying the classic Marxian self-destructive, self-survival twist of capitalism) heading towards peak data, meaning that there must be an endpoint of data extraction that it no longer makes economic reason to sustain the extraction of data. From this conclusion, the SMSM provides an outlet for this tension and further re-affirms social media as being part of the mixed model.

2. Subscription Services

For many, the idea of subscription services for social media began with Elon Musk’s recent implementation of a subscription service on X (formerly known as Twitter), “Twitter Blue”. However, many social media companies already had some form of subscription method prior to Musk’s development, or at least already had plans for one. Even Twitter stated their intentions of developing a subscription model for 2023, years before Musk’s takeover (McMillan 2020; Needleman 2021). In fact, “Twitter Blue” was rolled out in Canada and Australia in June 2021, at least a full year before Musk’s ownership of Twitter (Gupta 2021).

An early and (in)famous statement that social media could become a subscription service came from Mark Zuckerberg. In April 2018, Zuckerberg was brought to the US Senate and questioned about Facebook’s role in the 2016 US Presidential election. During his appearance, former Senator Orrin Grant Hatch asked Zuckerberg, “In 2010… you said back then that Facebook would always be free. Is that still your objective?” to which Zuckerberg responded, “Senator, yes. There will always be a version of Facebook [Meta] that is free”(ABC 2018). Clearly, he was implying that at some point Meta would have some form of a fee-based system. Also, Meta famously (and quietly) changed their original slogan: “It’s free and always will be” to “It’s quick and easy” (Asenjo 2019). A few years later, Meta announced that they were trialling a subscription service for Facebook and Instagram (Taylor 2023).

YouTube, owned by Google, created YouTube Premium in 2014. It was created as a Music service platform so that the users could pay a fee to avoid advertisement (similar to Spotify and Pandora services). It has since morphed into a service that can block advertisements on “millions of videos”, “download videos to watch offline”, and “play in the background: Keep playing videos when using other apps or when your screen is off” (‘YouTube Premium’). It has multiple payment schemes: a family plan, a regular plan, or a student plan available (YouTube). YouTube boasted in 2021 to hit the 50 million subscriber mark (YouTube 2021). Although this sounds like a significant amount, this only covers a small percentage of YouTube’s audience: “[YouTube] signed up 50 million subscribers across its suite of paid services, which include a music service and a premium video service. It convinced about 2.5% of its users to pay in order to watch videos in the background, skip ads and access all kinds of music playlists” (Shaw 2021). Later, in 2022, YouTube announced that it hit 80 million premium subscribers (Cohen 2022). This fits with a rise in subscription service outside of social media over recent years. Spotify, for example, went from 6 million paying-subscribers in 2013 to 68 million paying-subscribers by the end of 2015 (Arditi 2018, 308). By 2023, it had over 210 million subscribers (Porter 2023).

Reddit has had a subscription service since 2010. Originally called Reddit Gold, now Reddit Premium, paying members acquire certain benefits such as ad-free browsing, they get exclusive “Avtar gear”, and access to a specific Reddit premium member- only lounge, custom app icons and premium awards (Reddit). Until recently paying members would get “700 gold coins”, however, Reddit has recently changed this reward system and will be implementing something new (Peters 2023).

Snapchat also offers a subscription service: Snapchat+. Again, this service offers beneficial perks to its paying customers such as, one can pin their best friend at the top of their list, see if a person has watched their story multiple times via an eyes emoji attached to their friend profile of said viewer, and change the wallpaper of their chats. According to CNET: “More than 1 million people worldwide are paying for Snapchat's subscription service, the social media company said Monday [article published August 15th, 2022]” (Wong 2022). The service started in late June 2022 (Wong 2022), and recently, Snapchat announced that they now have over 2 million subscribers (Snapchat 2023). Musk might have brought subscription services into mainstream thought, but they clearly have been a factor in social media since the 2010s.

However, a question remains: since subscription services have been a part of social media for a considerable amount of time, why have they avoided critical engagement? The economic extraction of social media companies has mostly been conceived as data, and that the user’s data will be generated freely; stressing the importance of a fee as a point of economic extraction systemically falls outside of the dominant view of data as the source of economic extraction.

3. Blind Spot: Data VS. Fee

As noted in the introduction, generating profit for social media companies has generally been conceived as exploitation/extraction of user’s data that is repackaged and sold to advertisers (Allmer 2015; José van Dijck 2013; Flisfeder 2021; Fuchs 2011; Fuchs et al. 2012; Fuchs 2013, 2020, 2021c; Jarrett 2016, 2018, 2022; Srnicek 2017; Zuboff 2019). As we will see, there is variation of this thought, but the general principle of extracting data from the user is critical. From this conceptualisation, the fermentation of the blind spot of subscription services within social media becomes clear: paying for social media does not neatly fit with this understanding as the emphasis is placed on the fee and not on data.

One formulation of social media capitalising from users’ data is Christian Fuchs's application of Dallas Smythe’s idea of the audience commodity (Fuchs 2020, 2021a; Smythe 1977). Smythe’s original thought of the audience commodity is that individuals throughout the day, most of their time (except sleep) is commodified and sold to advertisers (Smythe 1977). Smythe opines that the audience, in essence, becomes part of the commodity process. When individuals are watching, listening, and engaging with a commercial media apparatus they are not simply consuming they are, in fact, labouring. How are they labouring? They are labouring because part of their engagement with media consumption finds itself attached to the commodifying process. This element of the commodifying process, is information that is returned to advertisers or marketing companies, which is why Smythe argues for the transformation of audience into the audience commodity (Fuchs 2021b, 162). The audience commodity, therefore, is creating value, in data or information, which is then sold on to advertisers. Audiences are unpaid for their worktime (watching the show etc.) and the data or information they create is exploited, and sold to advertisers and marketing which then return this to the individual as advertisements (Smythe 1977). Fuchs translates this logic into social media terms: “[social] media companies sell the audience as a commodity to advertisers” (Fuchs 2021b, 85). More specifically, the user becomes a commodity, via the user’s generated data (liking, commenting, posting, almost any online interaction). This data becomes layered with important economic capacities, unpaid labour, and exploitation, which then the corporate Internet platforms “offer the data commodity that is the result of Internet prosumption activity for sale to advertisers” (Fuchs 2015, 708).Therefore, Fuchs’ use of Smythe’s audience commodity concept can be brought to the 21^st century and social media users embody similar logic because they are exploited, their data is extracted and alienated from the products of the labour (every interactions online) in a similar manner as how Smythe developed. According to Fuchs, “[t]he time spent on corporate platforms is the value created by their unpaid digital labour” (Fuchs 2015, 708). These developments and economic logic lay the foundation for social media’s economic base; as Fuchs states: “Targeted advertising is social media platforms’ dominant capital accumulation model” (Fuchs 2021b, 86).

Nick Srnicek has advanced the idea of platform capitalism^{^[1]}. Srnicek centres data as the key development for 21st-century capitalism and platform capitalism. (Srnicek 2017). Data has become the key tool for multiple different aspects of digital capitalism: “Data extraction is becoming a key method of building a monopolistic platform and of siphoning off revenue from advertisers” (Srnicek 2017, 58). Within the platform capitalism approach, there are five types of platforms that are created: advertising platforms, cloud platforms, industrial platforms, product platforms and lean platforms (Srnicek 2017, 49). The relevant one for this discussion is the advertising platform. Srnicek emphasises the importance of data and states that platform companies are designed to be “an extractive apparatus for data” (Srnicek 2017, 48). Srnicek’s starting point for these platforms was the Dot Com Bust at the start of the 2000s, which caused many businesses to fail. This left space for a monopolistic rise for certain companies (Google, Facebook) (Srnicek 2017, 50-51). Another fallout from the bust was that venture capitalists shifted investment in the market to relay on steady revenue streams: this reliable revenue stream was advertisement (Srnicek 2017, 51). As will be discussed in more detail in the Zuboff section, this created an economic incentive to target advertising, which was the first major shift towards Surveillance Capitalism.

Srnicek distances himself from a more Marxian or Fuchs/Smythe interpretation of the role of labour (Srnicek 2017, 53-56). He fears that the role of labour becomes too totalising and that the claim everything becomes commodified leaves no social interactions outside of capitalism^{^[2]} (Srnicek 2017, 54). Srnicek argues that most of our social engagement is not “co-opted into a system of profit generation” and if they already were, why would companies develop “the extractive apparatus of platforms” (Srnicek 2017, 54)? Platform companies also rely on other forms of economic sources; for example, Google uses “information collected by sensors in the internet of things, corporate and government data […] and public and private surveillance […]” (Srnicek 2017, 54-55). Further, these interactions, for Srnicek, cannot be considered labour-producing since the role of digital labour does not fall under the same pressure of traditional, capitalistic labour such as “to rationalise the production process, to lower costs, to increase productivity, and so on” (Srnicek 2017, 55). And finally, if it were a new source of free labour it would open up a new area of exploited labour; whereas, in fact, the firms and platforms take value away from other areas of global capitalism (Srnicek 2017, 56). Therefore, for Srnicek it makes more sense to see what advertising platforms are extracting is not free labour, but “data as raw material” (Srnicek 2017, 56)^{^[3]}. Importantly, this data goes through a cleaning process (Srnicek 2017, 57), in that advertisers are not receiving bland, unorganised points of data. They have been laboured on in order for the advertisers to have a good picture of a particular user (Srnicek 2017, 57), increasing the rise and importance of surveillance of the user (again, more of this will be discussed in Zuboff’s section).

Yanis Varoufakis's most recent text, Techno Feudalism (Varoufakis 2023), develops the notion of cloud capital (Varoufakis 2023, 62)^{^[4]}. Cloud capital started as a basic sorting or aggerating of searches that could “guide us to the websites, friends, colleagues […] that we might like […] they [cloud developers] wrote algorithms capable of categorising us in clusters of internet users with similar search patterns and preferences” (Varoufakis 2023, 75)^{^[5]}. However, three key steps in the development of algorithms changed the course of the development of cloud capital. First, the process of reprograming algorithms’ objectives was salient, in that the algorithm could learn and then come up with further solutions. Second, algorithm hardware changed in that it could be more dynamic and “process a lot more information faster […] adopting layered network structures that allow for the interconnection of many different nodes” (Varoufakis 2023, 76). Third, and most importantly, reinforced-learning allows the algorithm to process its performance and improve to fit what the engineer programmes the algorithm to do (Varoufakis 2023, 76-77). Cloud capital now has the capacity to influence individuals on an unprecedented scale (Varoufakis 2023, 79-80). Furthermore, Varoufakis, echoes Fuchs’s thought that cloud capital is capitalising from individual free labour which directly benefits the owners of cloud capital (Varoufakis 2023, 78). Importantly, what sustains and makes cloud capital distinctive is that the model functions from the reproduction of free labour instead of wage labour (Varoufakis 2023, 80). Varoufakis also introduces the idea of cloud serfs (similar to prosumers) which in essence is anyone engaging with the cloud (Varoufakis 2023, 83-85) (engaging with the cloud, again, replicates similar dynamics already seen: uploading pictures, watching a YouTube video, liking a comment, sending a tweet, even driving a Tesla). Engaging in cloud work means replenishing cloud capital’s stock with constant interaction with the online world and the cloud^{^[6]}. Cloud serfs are integral to the function of the system as they provide the stock, or the data, to sustain the system. Reproducing billions of people as part of the free workforce is, according to Varoufakis, the most unique revolutionary part of cloud capital: “The true revolution cloud capital has inflicted on humanity is the conversion of billions of us into willing cloud serfs volunteering to labour for nothing to reproduce cloud capital for the benefit of its owners” (Varoufakis 2023, 85). Finally, as the title suggests, Varoufakis argues that we have shifted into a sort of neo-feudal era with the re-emergence of rent as the driving force: “It is this fundamental fact – that we have entered a socio-economic system powered not by profit but by rent – that demands a new term […] I can think of no better name than technofeudalism” (Varoufakis 2023, 132)^{^[7]}. Cloud capital engages with rent (or cloud fiefs to use Varoufakis’ term) insofar that, for example, Apple charges a 30% price for developers to use the Apple store (Varoufakis 2023, 125).

Fuchs, Srnicek, and Varoufakis may have conceptual, subtle, and theoretical disagreement with the role of labour, the totalising impact of capitalism, attention commodity and the (re)production of neo-feudalism vs neoliberalism. However, a point remains untouched in all three approaches: acquiring and selling users’ data with sophisticated surveillance, moulding and monitoring techniques with a capitalist/commodifying end (selling/renting the data to advertisers) is the dominant economic model for social media. This centring of data has allowed for critiques of subscription services of social media to fall by the wayside. Many paying for social media does not nicely fit into the freely accessible social media model. It systematically falls outside of critiques and criticism that emphasise data and not the user paying a fee.

Fuchs provides the solution and the problem of this tension in his text, Communication and Capitalism (Fuchs 2020). In this text, Fuchs provides a table of “[m]odels of capital accumulation in the capitalist communication industry” (Fuchs 2020, 135). He provides ten categories that list different sorts of models that impact different types of accumulation depending on the type of communication company. Fuchs notes that the models are not “mutually exclusive. There are communication corporations that accumulate capital by combining several models” (Fuchs 2020, 134). Fuchs, for example, cites Amazon which is part of the online retail model, whilst also being part of the subscription and streaming model, Amazon Prime (Fuchs 2020, 134-135). The final category of the table consists of the “mixed models”. As mentioned in the introduction, a company engaging with a mixed model approach means a company utilises multiple forms of capital accumulation, and they are not locked into a specific tactic. For example, Spotify is both advertisement-based, whilst also subscription-based (Fuchs 2020, 136). Fuchs notes that many companies engage in this tactic (Fuchs 2020, 136). Importantly, however, there is no mention of social media as having a crossover. Social media companies do not appear in the category of the mixed models or the subscription and streaming model, nor are they mentioned as having a crossover with subscription models in the proceeding text discussing the table (Fuchs 2020, 134-137)^{^[8]}. Fuchs provided the solution, insofar as it is now time to consider social media companies as being part of a mixed model and subscription approach. Fuchs has also outlined the problem: social media’s main method of economic extraction, extraction of user’s data and sold (or rented) to advertisers, excludes or heavily neglects the subscription side because paying for social media systematically cannot fit into the free access model^{^[9]}. Therefore, the subscription side of social media has been able to avoid critique and remain dormant.

The SMSM has remained an uncritiqued blind spot in many contemporary analyses of social media, due to, on the one hand, the novelty of these ideas being in the collective cultural consciousness. On the other hand, the emphasis placed on data as the main source of economic extraction has allowed for systematic neglect of subscription and oversight of the SMSM. A justification is also required for the emergence of subscription services. Subscription services and the SMSM are a structural response to a contradiction within surveillance capitalism, although it can be expanded out to the critiques given above, that contradiction being “peak data”. Here, this article adds to the many important developments that have caused problems for the surveillance model and argues that the SMSM is placed within a contradiction embedded in the surveillance capitalism and, in a more general sense, data-extracting models.

4. Cracks in The Surveillance Model

This section will discuss key aspects of Zuboff’s notion of surveillance capitalism^{^[10]} and her “no exit” situation and stress the contradiction of peak data that in turn justifies the emergence of the SMSM, whilst further cementing the coexistence of social media within a Mixed Model approach. It is worth noting that the other approaches discussed above, Srnicek’s notion of platform capitalism, Smythe/Fuchs’ concept of audience commodity and Varoufakis’ idea of cloud capital, can relate to this contradiction. However, due to the influential nature of Zuboff’s account and of the importance of the surveillance of the user for the economic function of social media (Fuchs 2021c, 140), I will focus on Zuboff’s account of surveillance capitalism.

The major aim of surveillance capitalism (similar to Varoufakis’ fears) is to act as a totalising force to extract and commodify “every aspect of every human’s experience” (Zuboff 2019, 9). As noted, this totalising aim arose out of the Dot Com Bust, whereby the crash caused a shift in the bases of the digital market: economic extraction towards an advertising model (Zuboff 2019, 63-98; Srnicek 2017, 50-53). As noted by Srnicek, Google was the first major company to thrust surveillance capitalism onto the world (Zuboff 2019, 65). Importantly, extraction of users’ data was already occurring; Zuboff calls this extraction, “The Behavioural Value Reinvestment Cycle” whereby Google would use users’ data to further improve their features (Zuboff 2019, 70). However, after the Dot Com Bust, Google shifted data extraction to fit the business demands of an advertising model: “The Discovery of Behavioural Surplus” (Zuboff 2019, 97). The difference between the two is that instead of improving features for the users (although improvement for the user’s experience still occurs under the Behavioural Surplus Model), the extracted data was repackaged as a source of behavioural surplus which was sold to advertisers to meet the advertisers’ end (Zuboff 2019, 78). These slight changes set off a domino effect for surveillance of individuals as profit (Zuboff 2019, 52), whereby the development of surveillance techniques to extract data led to a loss of privacy, to which privacy or loss of privacy became a sort of “price one must pay for the abundant rewards […]” (Zuboff 2019, 52; Mosco 2014, 138). Establishing an economic incentive: invasive, totalising, surveying tools (algorithm-learning, tracking of users, cookies, listening into people’s conversation (Amazon’s Alexa, for example) etc.) were developed in order to extract as much data as possible. This was done to “nudge, coax, tune, and herd behaviour toward profitable outcomes” (Zuboff 2019, 8).

Surveillance capitalism is also not ahistorical and was created during a neoliberal hegemony which helped ferment the logic of these companies (Zuboff 2019, 107-112). For example, Fuchs notes: “The collection, storage, control, and analysis of “big data” stands not just in the context of the surveillance-industry complex but also in the context of neoliberalism” (Fuchs 2021c, 53). Furthermore, the terror attack of 9/11 played an influential role in changing the world’s view on surveillance (Zuboff 2019, 112-121; Mosco 2014, 151-152). For example, Zuboff highlights how in 2000 the US Federal Trade Commission debated the invasion of privacy of the developing surveillance companies and they came to the conclusion that self-regulation would not be enough and that privacy legalisation would be needed (Zuboff 2019, 111). However, after 9/11 the debate stopped, and surveillance capacities expanded. Post-9/11 helped to create a sense of surveillance exceptionalism. This surveillance exceptionalism was expanded by the mutual cooperation between surveillance companies and governments (Zuboff 2019, 112-121). Plus, the expansion of legislation such has the Patriot Act (in the US) helped to foster this ethos and reduce debate about the dangers of these companies (Zuboff 2019, 114)[11]. In her final chapter, The Right to Sanctuary, Zuboff elaborates on the “no exit” situation of surveillance capitalism. She argues that the all-encompassing and intrusive nature of surveillance capitalism generates a need for and importance of sanctuary from the Big Other. For Zuboff, a “no exit” is an essentialist situation whereby behavioural surplus is transformed from the individual into revenue, the individual’s autonomy has been transfixed to a profit end and removal of self-moral judgement for an inescapable control of digital companies (Zuboff 2019, 488). In a sense, a “no exit” situation is when surveillance capitalism has won and there is no real way to escape from the dominion of surveillance companies. Hence, her stress on providing sanctuary and protection from these companies and the intrusive, totalising and controlling pressure of the Big Other (Zuboff 2019: 488). To summarise, the discovery of surplus behaviour, technological developments, the erosion of privacy, tracking of users, surveillance industry, economic incentive, neoliberal ideology, neoliberal (lack of) regulation and historical events, laid the foundational blocks for the critical logic of for the accumulating data. However, surveillance capitalism is starting to show cracks.

To focus on Meta, Dan Milmo, the technology editor for the Guardian, highlights the economic glue holding Zuckerberg’s business together: “advertising-based model accounted for 98% of Meta’s $116.6bn in revenues last year, but it faces problems” (Milmo 2023; emphasis added)^{^[12]}. The three problems that Meta faces are Apple’s changed privacy policy, wider legal and regulatory alterations, and a general economic stagnation of social media companies (Bobrowsky 2022; Milmo 2023b; Newman 2022). Apple changed their data collection policy allowing users to opt out of being tracked which has made it more difficult to collect data (Bobrowsky 2022; Newman 2022). Consequently, Meta stated that this change in Apple’s policy, “would cost it $10 billion, on account of how much more difficult it's become to target ads to iPhone users” (Zaveri 2023). Further, a court ruling stated that Meta broke EU data laws with personalised ads (Milmo 2023b). Meta was fined £343 million by Ireland’s Data Protection Commission after “the EU’s data authority rejected the company’s argument that users agree to receive ads based on their personal data when they enter into a ‘contract’ with its social media platforms via the terms and conditions they sign.” (Milmo 2023b). After this decision, Max Schrems the chair of Noyb (the group that brought this case against Meta) said “Facebook and Instagram users in the EU would now need to be asked whether they wanted their data to be used for ads” (Milmo 2023b). Milmo further quoted Schrems: “This is a huge blow to Meta’s profits in the EU, … People now need to be asked if they want their data to be used for ads or not. They must have a ‘yes or no’ option and can change their mind at any time” (Milmo 2023b). Zuboff stresses Surveillance Capitalism can be curtailed via regulations/laws. For example, Zuboff stated in an online debate/discussion that: “[…] the reason that these companies [Surveillance Capitalists] can drive for more, more, more is that in the history of capitalism, there is only one thing that curbs the drive for more, more, more […] that one thing is law” (UCL 2021). She placed hopes in the EU’s General Data Protection Regulation as an example of curtailing these companies (Zuboff 2019, 481). Meta themselves have responded that they are introducing a subscription add-free experience for their EU members thanks to the role of new legislation (Meta 2023). Furthermore, Milmo highlights that there is a “general economic weakness, to which any advertising dependent social media or search company is susceptible”, and other companies have felt this economic weakness: “In its most recent quarterly results, Meta reported a 4% drop in revenue while other platforms, including Snap, Google and Elon Musk’s Twitter, have also warned of a weak advertising environment” (Milmo 2023a). Finally, as another extension of regulation, there have been calls to introduce some form of an advertising tax (Fuchs 2021c, 352; 2018).

To give another example, at the announcement of Musk’s subscription service, Stephen King tweeted: “$20 a month to keep my blue check? Fuck that, they should pay me. If that gets instituted, I’m gone like Enron.” (King 2022). To which Musk responded: “We need to pay the bills somehow! Twitter cannot rely entirely on advertisers. How about $8?” (Musk 2022). Of course, Musk could be overestimating the importance of the new project. Importantly, as Musk’s takeover of Twitter has not been a success for the billionaire, shifting to Twitter Blue could be seen as an attempt to save face^{^[13]} (Milmo 2023c). However, this shift towards a subscription service highlights a shift in the industry, as Jake Rudnitsky writes in Bloomberg, “Meta Platforms Inc. is the latest social networking company to ask users to pay for account verification, joining attempts from competitors including Twitter Inc. to diversify away from ad revenue” (Rudnitsky 2023). In an MSNBC article, the author cites a Bank of America report that states that Meta could have 12 million paying users by early 2024, which is “worth about $1.7 billion in revenue” (Munk 2023). Another problem the data-collection model has caused is destabilising effects on democracies, as seen with the Cambridge Analytica scandal (Fuchs 2021, 150-153). In a sense, Cambridge Analytica used targeting data and targeting advertisement “for targeting voters with political ads and fake news in elections” (Fuchs 2021c, 151). Furthermore, Fuchs cites a Guardian article (Cadwalladr 2020) that highlights that over 100,000 Cambridge Analytica documents which interfered with 68 elections (Fuchs 2021c, 151). Furthermore, Facebook was fined $5 billion by the Federal Trade Commission for “violating users’ privacy in the context of the Cambridge Analytica scandal” (Fuchs 2021c, 151).

Although these examples mostly focus on regulation in the EU and the US, a particular social media company, and social media exist outside the EU and the US, and their generative capacity for profit exists outside of the EU and US jurisdiction. It is still a useful blueprint to show cracks in the wider advertisement base of the surveillance model. However, I want to add a rarely acknowledged contradiction to the already forming problems of the surveillance model, advertising model or to use Srnicek’s phrase platforms that are “an extractive apparatus for data”: the collection of data may swallow itself whole and lead to “peak data”. This classical Marxist twist suggests that surveillance capitalism’s contradiction provides a point of self-destruction, which further codifies itself and stresses the necessity of social media and the SMSM as part of the mixed model category.

5. The Marxian Twist on Peak Data

It is well known in the Marxist tradition that capitalism must continually destroy itself in order to recreate itself anew. As Slavoj Žižek states: “capitalism is capable of transforming its limit, its very impotence, in the source of its power – the more it ‘putrefies’, the more its immanent contradiction is aggravated the more it must revolutionize itself to survive” (Žižek 2008, 53). Rosa Luxemburg states similar thoughts: “Like them [money, commodities and capital], it [credit] is an indispensable gear in the mechanism of capitalist economy, and at the same time, an instrument of destruction, since it aggravates the internal contradictions of capitalism” (Luxemburg 2008, 70). Elsewhere, Luxemburg writes: “Crises therefore appear to be the instrument of rekindling the fire of capitalist development” (Luxemburg 2008, 71). Similarly, Marx stated in his third volume of Capital: “The true barrier to capitalist production is capital itself… (Marx et al. 1946, 346); and, famously, in the Communist Manifesto: “The weapons with which the bourgeoisie felled feudalism to the ground are now turned against the bourgeoisie itself. But not only has the bourgeoisie forged the weapon that brings death to itself…”(Marx and McLellan 2000, 250). Capital, to summarise, is its own negation.

Applying these ideas to surveillance capitalism, or to platforms that extract data, there arises a contradiction: the aim of these entities is to acquire data, whether that be all human experiences, for adversities, for the audience commodity, for the platform or for resupplying the Cloud. However, what happens to these entities if they extract all data, or at least collect a sufficient amount of data whereby anymore data does not unearth any more relevant information for advertisers, the Cloud, or for the platform? The totalising, commodifying, alienating, and monitoring of human expression not only causes a problem for individuals but also for the model of surveillance capitalism itself; in essence, what happens when the treasure trove of data has been exhausted? Extracting data from a user must have a certain endpoint, technologically or conceptually, which is driving the extracting data model. How can social media work its way out of this tension?

Until now, this contradiction, which drives social media, is rarely considered. Geert Lovink, in his introduction to his Sad by Design (Lovink 2019), does bring up this contradiction with the idea of peak data which is a “theoretical point when the maximum rate of data extraction has been attained” (Lovink 2019, 7). Once this point has been reached, the data system collapses and: “The goldmine of data suddenly becomes digital garbage” (Lovink 2019, 8). Lovink advances on this stance and argues that digital companies are progressing towards a reduction of data collection, by reducing the use of users (Lovink 2019, 8-9). For example, YouTube reminds users to ‘take a break from the app’ (Lovink 2019, 8). iPhone themselves also have instilled this feature in a manner of self-regulation with time limits for certain apps. This is then repackaged as companies being responsible and looking out for the user (Lovink 2019, 9).

While Lovink's unique analysis of “peak data” is useful for highlighting the contradiction, a better option for dealing with this tension is the assertion of the SMSM. Lovink’s analysis is still working under the blind spot, and he does not consider the subscription services as a way out of the peak data tension. For example, YouTube Premium offers a service that paying customers would not see advertisements while engaging on their videos (excluding, of course, the people uploading the videos may include their own add/sponsorship). The logic for advertisers to buy/rent advertisement space (or our attention to using Fuchs/Smythe’s logic) – becomes problematic if there is no means for advertisers to advertise. Or to use Meta’s own words: “If a person chooses to subscribe, they won’t see ads and we will not process their information for personalised advertising” (Meta 2023).There may still be advertisements, but the targeting of personalised advertisement, in Meta’s words, seems to be changing. To be more specific, the SMSM helps reduce the tension of peak data whereby, data will still be collected but the burden of dependence is shifted from mostly the advertisement/data-collection model to a mixture of subscriptions and data collection (exactly what Meta is doing). To use a psychoanalytic phrase, the SMSM acts as a sort of necessary surplus that is not subsumed by the data-collection model but helps to sustain the model itself.

Importantly, the SMSM, in no sense, suggests that these platforms/companies will drastically shift and erode entirely from a surveillance model/data-collection model. There are much larger political, economic, legal, social, ideological, and historical reasons that fix the surveillance/data-collection model. Something of this radical implication would most likely be a much slower and cumbersome process. However, to summarise what is being stated: the SMSM asserts that the role of subscription services and argues that they have been overlooked. This has created a blind spot, whereby, social media critiques have overlooked the impact of subscription services which now defines more clearly that social media is part of a Mixed Model approach, not only for the empirical base but also for the final point of the argument, namely, that the SMSM provides an outlet for the contradiction of “peak data” whilst affixing social media as part of a Mixed Model.

6. Conclusion

This article asserts the SMSM as an essential step in conceptualising an important area of social media. The SMSM aids the explanation of a blind spot of contemporary critiques of social media, the subscription services. Second, with the developing cracks in the surveillance model, the SMSM embraces the contradiction of “peak data” that sustains multiple interpretations of the economic base of social media, collection of user’s data, and allows for an outlet of this contradiction, which further reconstitutes social media as part of the mixed model.

Until the end of 2022, subscription services on social media remained relatively unimportant within cultural consciousness, which potentially, allowed for a lack of engagement. Musk’s development, rash character and handling of Twitter Blue helped to stimulate subscription services within cultural consciousness. However, within academia they remain relatively obscure. The integral reason that the SMSM remained ignored is that it systematically falls outside of contemporary critiques because the acquisition of user data is not stressed. Data is the core of social media’s economic survival; to reuse Srnicek’s phrase, most critiques begin with considering social media as an extractive apparatus for data. However, within the SMSM things are slightly different. Of course, the extracting data will continue, but within the SMSM a stressor is also placed on the user paying for social media. Because the SMSM does not necessarily emphasise data, it has been neglected by criticisms. The saliency of SMSM also transforms and cements social media as part of the mixed model approach. Importantly, developments have led to problems with the status quo of social media and unhinged access to data. Legal developments, destabilising impacts on democracies, scandals, violations of EU law, and fines from the US government have caused important regulations both internally and externally and a re-evaluation of social media’s surveillance and data extractive capacities (Meta already seems to be responding to these changes). However, to further cement social media as part of a Mixed Model is to acknowledge the tension of “peak data”. To take surveillance capitalism/advertising models/data-collection capital accumulation models to their logical end, there must be an endpoint whereby sufficient data is collected and the act of collecting data becomes less economically sensible. Therefore, the SMSM is a method for social media to deal with the contradiction of “peak data”, and further establish social media as a Mixed Model.

To end, viewing social media as part of a mixed model with an important and active SMSM unearths new questions and tension that demand future research and theoretical rumination. Returning to Varoufakis’s thought, if one considers the subscription fee as a source of rent, does this change the logic of social media? Is it no longer reproducing neo-liberal logic and is it shifting towards a neo-feudal logic? Is it now time to consider a sixth form of platform capitalism, the subscription platform? From an audience commodity perspective, do members of the SMSM act more like consumers or still fixed as prosumers, or both? Democratically, what dangers or problems will a distinctively hierarchical social media contain? Would this mixed model approach, however, emphasise a rise of a freer version and more democratically organised, Mastodon for example? But, before these questions (and many more) can be addressed, the first step is to acknowledge the to the role of subscriptions services on social media and retain social media as part of the mixed model.

References

ABC, News. 2018. Mark Zuckerberg: 'There Will Always be a Version of Facebook that Is Free'. Accessed January 1, 2023.

https://www.youtube.com/watch?v=kFWO6XnYoVU.

Allen, Bobby. 2022. Elon Musk Relaunches Twitter Blue, the Subscription that Includes Checkmarks. NPR. Accessed January 2, 2024.

https://www.npr.org/2022/12/12/1139619598/twitter-blue-relaunch.

Allmer, Thomas. 2015. Critical Theory and Social Media: Between Emancipation and Commodification. Critical Theory and Social Media. Abingdon, Oxon: Routledge.

Arditi, David. 2018. Digital Subscriptions: The Unending Consumption of Music in the Digital Era. Popular Music and Society 41 (3): 302-318.

https://doi.org/https://doi.org/10.1080/03007766.2016.1264101.

Asenjo, Qayyah Moynihan & Alba. 2019. Facebook Quietly Ditched the 'It’s Free and Always Will Be' Slogan from Its Homepage. Business Insider. Accessed March 3, 2023.

https://www.businessinsider.com/facebook-changes-free-and-always-will-be-slogan-on-homepage-2019-8?r=US&IR=T.

Bobrowsky, Meghan. 2022. Facebook Feels $10 Billion Sting from Apple’s Privacy Push. The Wall Street Jounral. Accessed Janurary 5, 2023.

https://www.wsj.com/articles/facebook-feels-10-billion-sting-from-apples-privacy-push-11643898139.

Cadwalladr, Carole. 2020. Fresh Cambridge Analytica Leak ‘Shows Global Manipulation Is Out of Control’. The Guardian. Accessed January 03, 2024.

https://www.theguardian.com/uk-news/2020/jan/04/cambridge-analytica-data-leak-global-election-manipulation.

Cohen, Lyor. 2022. We Hit 80 Million?!. YouTube. Accessed Feburary 7, 2023.

https://blog.youtube/news-and-events/youtube-music-premium-80-million-lyor-cohen/.

FitzGerald, James. 2022. Twitter Confirms Fee for Blue-Tick Verification after Musk Takeover. BBC. Accessed January 2, 2024.

https://www.bbc.co.uk/news/technology-63530872.

Flisfeder, Matthew. 2021. Algorithmic Desire: toward a New Structuralist Theory of Social Media. Evanston: Northwestern University Press.

Fuchs, Christian. 2011. Web 2.0, Prosumption, and Surveillance. Surveillance & Society 8 (3): 288-309.

Fuchs, Christian. 2013. Political Economy and Surveillance Theory. Critical Sociology 39 (5): 671-687.

Fuchs, Christian. 2015. Dallas Smythe Today–The Audience Commodity, the Digital Labour Debate, Marxist political Economy and Critical Theory. Prolegomena to a Digital Labour Theory of Value. In Marx and the Political Economy of the Media, 522-599. Leiden: Brill.

Fuchs, Christian. 2018. The Online Advertising Tax as the Foundation of a Public Service Internet. London: University of Westminster Press.

Fuchs, Christian. 2020. Communication and Capitalism: a Critical Theory. Critical Digital and Social Media Studies. London: University of Westminster Press.

Fuchs, Christian. 2021a. Foundations of Critical Theory: Media, Communication and Society Volume Two. New York: Routledge.

Fuchs, Christian. 2021b. Marxist Humanism and Communication Theory: Media, Communication and Society Volume One. New York: Routledge.

Fuchs, Christian. 2021c. Social Media: a Critical Introduction. Los Angeles: SAGE. Third edition

Fuchs, Christian, Kees Boersma, Anders Albrechtslund, and Marisol Sandoval. 2012. Internet and Surveillance: The Challenges of Web 2.0 and Social Media. Vol. 16. New York: Routledge

Gupta, Sara Beykpour and Smita. 2021. "Introducing Twitter Blue - Twitter’s First-Ever Subscription Offering." Twitter. Accessed Feburary 16, 2023. https://blog.twitter.com/en_us/topics/company/2021/introducing-twitter-blue.

Jarrett, Kylie. 2016. Feminism, Labour and Digital Media: the Digital Housewife. New York: Routledge.

Jarrett, Kylie. 2018. Exploitation, Alienation, and Liberation: Interpreting the Political Economy of Digital Writing. In The Routledge Handbook of Digital Writing and Rhetoric, 423-432. New York: Routledge.

Jarrett, Kylie. 2022. Digital Labor. Cambridge: Polity Press.

King, Stephen (StephenKing). 2022. "$20 a Month to Keep My Blue Check? Fuck That, They Should Pay Me. If That Gets Instituted, I’m Gone Like Enron." Twitter, 31 October, 2022, 11:23 am. https://twitter.com/StephenKing/status/1587042605627490304.

Lovink, Geert. 2019. Sad by Design: on Platform Nihilism. London: Pluto Press.

Luxemburg, Rosa. 2008. The Essential Rosa Luxemburg: Reform or Revolution and the Mass Strike. Edited by Helen Scott. Chicago: Haymarket.

Marx, Karl, Friedrich Engels, Samuel Moore, and Edward B. Aveling. 1946. Capital: a Critical Analysis of Capitalist Production. London: Allen & Unwin.

Marx, Karl, and David McLellan. 2000. Selected Writings. New York: Oxford University Press. Second edition

McMillan, Robert. 2020. Twitter Signals Interest in Developing Subscription Service. The Wall Steet Journal. Accessed Janurary 21, 2023.

https://www.wsj.com/articles/twitter-signals-interest-in-developing-subscription-service-11594245176?mod=article_inline.

Meta. 2023. Facebook and Instagram to Offer Subscription for No Ads in Europe. Last Modified December 4, 2023. Accessed January 9, 2024.

https://about.fb.com/news/2023/10/facebook-and-instagram-to-offer-subscription-for-no-ads-in-europe/.

Milmo, Dan. 2023a. Does Paid-for Facebook and Instagram Signal End of Free-Access Orthodoxy?. The Guardian. Accessed Janurary 12, 2023.

https://www.theguardian.com/technology/2023/feb/20/facebook-instagram-paid-for-signal-free-access-meta.

Milmo, Dan. 2023b. Meta Dealt Blow by EU Ruling that Could Result in Data Use ‘Opt-in’. The Guardian. Accessed April 12, 2023.

https://www.theguardian.com/technology/2023/jan/04/meta-dealt-blow-eu-ruling-data-opt-in-facebook-instagram-ads.

Milmo, Dan. 2023c. ‘Musk Destroyed All that’: Twitter’s Business Is Flailing after a Year of Elon. The Guardian. Accessed October 30, 2023.

https://www.theguardian.com/technology/2023/oct/27/elon-musk-x-twitter-takeover-revenue-users-advertising.

Mosco, Vincent. 2014. To the Cloud: Big Data in a Turbulent World. Boulder: Paradigm Publishers.

Munk, Cheryl Winokur. 2023. The Big Security Risks behind Meta, Twitter Verified Identity Subscriptions. CNBC. Accessed May 18, 2023.

https://www.cnbc.com/2023/02/23/biggest-benefits-risks-in-meta-twitter-verification-subscriptions.html.

Musk, Elon (elonmusk). 2022. We Need to Pay the Bills Somehow! Twitter Cannot Rely Entirely on Advertisers. How about $8? Twitter. November 1, 2022, 5:16 am. https://twitter.com/elonmusk/status/1587312517679878144?lang=en.

Needleman, Sarah E. 2021. Twitter to Launch Subscription Service Super Follows, Aims to Double Revenue by 2023. The Wall Street Jounral. Accessed Janurary 4, 2023.

https://www.wsj.com/articles/twitter-aims-to-double-revenue-by-2023-11614266567?mod=article_inline.

Newman, Daniel. 2022. Apple, Meta and the $10 Billion Impact of Privacy Changes. Forbes. Accessed May 11, 2023.

https://www.forbes.com/sites/danielnewman/2022/02/10/apple-meta-and-the-ten-billion-dollar-impact-of-privacy-changes/?sh=337b522572ae.

Peters, Jay. 2023. Reddit Is Getting Rid of its Gold Awards System. The Verge. Accessed March 19, 2023.

https://www.theverge.com/2023/7/13/23794403/reddit-gold-awards-coins-sunset.

Porter, Jon. 2023. Spotify Passes 500 Million Monthly Active Users for the First Time. The Verge. Accessed April 14, 2023

https://www.theverge.com/2023/4/25/23695790/spotify-earnings-q1-2023-monthly-active-users-515-million.

Reddit. Reddit Premium. Accessed May 16, 2023.

https://www.reddit.com/premium.

Rudnitsky, Jake. 2023. Meta Asks Users to Pay for Verification Following Twitter Move. Bloomberg. Accessed May 10, 2023.

https://www.bloomberg.com/news/articles/2023-02-20/meta-asks-users-to-pay-for-verification-following-twitter-move?leadSource=uverify%20wall.

Shaw, Lucas. 2021. YouTube is Free. But 50 Million People Pay for It Anyway. Bloomberg. Accessed March 11, 2023.

https://www.bloomberg.com/news/newsletters/2021-09-05/youtube-is-free-but-50-million-people-pay-for-it-anyways.

Smythe, Dallas W. 1977. Communications: Blindspot of Western Marxism. CTheory 1 (3): 1-27.

Snapchat. 2023. Snapchat+ Passes 2 Million Subscribers. Accessed May 7, 2023. https://newsroom.snap.com/en-GB/snapchatplus-twomillion.

Srnicek, Nick. 2017. Platform Capitalism. Cambridge: Polity Press.

Taylor, Josh. 2023. Facebook and Instagram to Get Paid Verification as Twitter Charges for Two-Factor SMS Authentication. The Guardian. Accessed June 11, 2023.

https://www.theguardian.com/technology/2023/feb/20/facebook-instagram-paid-meta-verification-twitter-charge-2fa-two-factor-authentication-via-sms.

Telegrapgh, The. 2023. Musk Tells Boycotting Advertisers to ‘Go F--- yourself’. The Telegrapgh. Accessed January 02, 2024.

https://www.youtube.com/watch?v=RK91Ji6GCZ8.

UCL. 2021. Digital Feudalism: The Future of Data Capitalism. UCL Insitute for Innovation and Public Purpose.

https://www.youtube.com/watch?v=fKgPQSa1_0o&t=3891s.

van Dijck, José. 2013. The Culture of Connectivity: a Critical History of Social Media. Oxford: Oxford University Press.

van Dijck, José, Thomas Poell, and Martijn de Waal. 2018. The Platform Society: Public Values in a Connective World. New York: Oxford University Press.

Varoufakis, Yanis. 2023. Techno Feudalism - What Killed Capitalism. Elcograf: The Bodley Head.

Wong, Queenie. 2022. Snapchat Reveals How Many People Pay for Its Subscription Service. CNET. Accessed May 10, 2023. https://www.cnet.com/news/social-media/snapchat-reveals-how-many-people-pay-for-its-subscription-service/.

Yaffe-Bellany, David. 2022. Twitter Begins Offering $7.99-a-Month Verification Subscriptions. The New York Times. Accessed January 2, 2024.

https://www.nytimes.com/2022/11/05/business/twitter-verification-check-marks.html.

YouTube. YouTube Premium. Accessed June 11, 2023.

https://www.youtube.com/premium.

YouTube. 2021. 50 million. Accessed April 11, 2023.

https://blog.youtube/news-and-events/50-million/.

Zaveri, Paayal. 2023. Meta's Cold War with Apple over Privacy is Coming to a Critical Moment. Business Insider. Accessed May 11, 2023.

https://www.businessinsider.com/meta-and-apple-user-privacy-dispute-ios-advertising-2023-2?r=US&IR=T#:~:text=In%202021%2C%20Apple%20changed%20its,target%20ads%20to%20iPhone%20users.

Žižek, Slavoj. 2008. The Sublime Object of Ideology. London: Verso.

Zuboff, Shoshana. 2019. The Age of Surveillance Capitalism: the Fight for a Human Future at the New Frontier of Power. London: Profile Books.

About the Author

Paul Geyer

Paul Geyer is a PhD student of Political Theory at the University of Leeds. His research focuses on social media, Political Theory, Critical Theory, and ideology. His current research project focuses on subscription services within social media and investigates how they are changing one's subjectivity and ideological relationship to capitalism, whilst also advocating for a neo-feudal understanding of the subscription side of social media.

[1] See The Platform Society for another useful description of the concept of Platform society/capitalism (José van Dijck, Poell, and Waal 2018).

[2] Kylie Jarrett’s concept of the digital housewife (Jarrett 2016, 2018) offers a useful way of responding to Srnicek’s fears by showing that a user can be both inside and outside of capitalist production but still be reproducing for a capitalist end. Whether that be subjectivity reproducing a subject to fit a capitalist end or from an economic standpoint that social interaction is nonetheless still commodified on social media even if the social interaction is not.

[3] A fair question, which many may respond to Srnicek with, is where does the raw material come from? Kylie Jarrett provides a useful analogue of considering social media as an assembly line with certain workers (engineers, software developers) as paid workers; whereas social media users who supply the raw material as unpaid workers (Jarrett 2016, 150). Srnicek’s perspective overlooks this thought experiment.

[4] Another useful earlier example of this dynamic is Vincent Mosco exanimation of Cloud computing in his To the Cloud – Big Data in a Turbulent World (Mosco 2014)

[5] As we shall see, this idea is very similar to Zuboff’s idea of “The Behavioural Value Reinvestment Cycle”.

[6] Similar to Fuchs’ user and the impact of the notion of the audience commodity.

[7] For more accounts of this concept see, (Arditi 2023; Company 2020; Dean 2020; Kotkin 2020; Waters 2020).

[8] A similar table can be found in another text (Fuchs 2021c, 81).

[9] Particularly, when one considers YouTube Premium or Twitter Blue as removing at least half of its advertisements from the paying user’s experience.

[10] Zuboff is not the only person to discuss Surveillance Capitalism. (see other interpretations of surveillance capitalism: (José van Dijck 2013; Fuchs 2011; Fuchs et al. 2012; Fuchs 2013; Mosco 2014).

[11] Europe and the EU also expanded similar regulations (Zuboff 2019, 114)

[12] Srnicek echoes these numbers: “Today Google and Facebook remain almost entirely dependent on them [advertisers]: in the first quarter of 2016, 89.0 per cent of Google’s and 96.6 per cent of Facebook’s revenues came from advertisers” (Srnicek 2017, 53).

[13] Particularly with the growing tension between Musk and advertisers. In a viral clip, Musk told advertisers who are boycotting X, to go “Fuck yourself” (Telegrapgh 2023).