NETFLIX – THE CURATION OF TASTE AND THE BUSINESS OF DIVERSIFICATION

Netflix is considered as a global business invested in strategies of diversification, localisation and personalisation in light of several discourses about the streaming service. One presents Netflix as an evil corporation encouraging binge-watching and reducing individuals to data. A utopian discourse proclaims the democratising potential of digital media technologies, including Netflix’s claims about its personalised, on-demand service. An industry discourse laments Netflix’s disruption of the film and television business. Finally, a scholarly discourse maps the political economy and cultural impact of Netflix. Each discourse attaches a particular cultural value to Netflix. Some offer ‘antidotes’, including the niche streamers, with their ‘curated’ collections of specialised content. Both types of streamer are in fact gatekeepers regulating access to cultural experiences and promoting particular ideas of taste and diversity. Netflix’s strategies of customisation and glocalisation, and its activities in the Middle East and North Africa, demonstrate in the end that diversity is good for business.


INTRODUCTION
Netflix is now truly a global media corporation, available in most countries of the world, with the notable exceptions of China, the Crimea, North Korea, and Syria, and worth some US$220 billion (CBS/AP, 2021). According to one estimate, with 208 million subscribers in more than 190 countries (Netflix, 2021a), "Netflix commands 19% of the global digital streaming market", and "12.6% of all downstream internet traffic worldwide" (Cook, 2021). Its catalogue is vast, with that same report suggesting 15,000 titles across all its sites (Cook 2021). That catalogue is, however, also limited in various ways, not least in its focus on scripted, professionally-produced TV series, documentaries and feature films, and increasingly in unscripted and reality formats, albeit across "a wide variety of genres and languages" (Netflix, 2021a). Although it has at times ventured into other areas, Netflix still proclaimed in 2021 that it was "not a generic 'video' company that streams all types of video such as news, user-generated, live sports, porn, music video, and gaming" (Netflix, 2021b). As a subscriberfunded service, it is unlike linear, advertising-supported television, not least in the way that it targets sensibilities, or taste clusters, rather than demographically defined audiences (Lotz, 2020). It is also unlike other internet-based platforms such as Facebook and YouTube, in that it is not a social medium, it is not open-access, and it does not depend on collaborative, user-generated material (Lobato, 2019;Lotz, 2017).
This article considers the ways in which Netflix as a major global business is invested in strategies of diversification, localisation and personalisation, and what this investment means in practice. I am interested in the relationships between the pronouncements the corporation makes about diversity and inclusion; its localised approach to how it selects, funds and presents the content available on its platform; and its commitment to a personalised, on-demand experience for its subscribers. To what extent is the industrial logic of customisation rather than standardisation consonant with diversity and inclusion policies? To what extent is the impression of unlimited abundance and endless choice that Netflix creates consonant with a policy of accessibility?
In exploring such questions, I engage with a range of discourses about Netflix. At one end of the spectrum is a discourse that presents Netflix as an evil neo-liberal corporation that hooks us into a binge-watching diet, reducing individuals to so much data, and cutting them off from the communal experience of the film theatre. At the other end are discourses that express a more utopian vision, whether proclaiming the democratising potential of digital media technologies, or the consumer-pleasing machinations of the on-demand culture. One of those utopian discourses consists of the proclamations that Netflix itself makes about its operations. For instance, "members can watch as much as they want, anytime, anywhere, on any internet-connected screen" (Netflix, 2021c), while "we are constantly working to make it easier for members to access and enjoy something they will love, regardless of their location or mood" (Netflix, 2021d).
Somewhere between the dystopian and the utopian discourses about Netflix is the discourse that responds with dismay to the disruptive impact that Netflix has had on the established film and television industries. This discourse is matched by more measured scholarly accounts of the political economy and socio-cultural impact of Netflix, notably in the work of McDonald and Smith-Rowsey (2016), Lotz (2017Lotz ( , 2020, Johnson (2019) and Lobato (2019), and others cited elsewhere in this paper.
Each of these discourses about Netflix attaches a particular cultural value to the global operations of Netflix -and some offer 'antidotes' to what they see as the poisonous nature of the business. I am interested in particular in the way niche streamers attempt to counteract what they see as Netflix's approach to global entertainment with their carefully curated collections of specialised content. I argue that both types of streamer are in the end gatekeepers that regulate access to cultural experiences and promote particular ideas of cultural value, taste and diversity. I also argue, however, that Netflix's strategies of customisation and glocalisation, and especially its activities in the Arabic-speaking countries of the Middle East and North Africa, demonstrate that diversity is good for business. "FREEDOM AND FLEXIBILITY": NETFLIX'S LIMITLESS OFFER Netflix is primarily a distributor of audio-visual content, operating a platform that provides story experiences for global audiences, enabling those audiences to select which audio-visual content to watch, when, how and under what conditions. As a business that generates income through subscriptions, and that competes with other content providers, and with other ways in which consumers spend their leisure time and earnings, Netflix is a content-hungry platform. It is constantly seeking to grow its global audiences, and to maintain those audiences it has already captured, by bringing them new material that caters to their varying tastes.
Their strategy now of course includes developing their own content with creators around the world. This business, and their online delivery system, has emerged in a particular cultural and industrial context, which brings together existing forms of media entertainment, particular developments in software and digital databasing, a membership subscription model, and a complex user interface.
Their content acquisition strategy has resulted in the creation of its huge and constantly changing online catalogue or database. Their content delivery strategy is organised around gathering and analysing huge amounts of data about its users, their viewing decisions and their preferences, and feeding that into its algorithmic recommender system to guide consumer choice (Leonard, 2013). Its 'unique selling proposition' is thus based on the size of its catalogue, enabling subscribers to "watch unlimited movies and TV shows", and a software system that enables "Freedom and Flexibility: Consumers can watch content on demand, on any screen, and the experience is personalized to individual tastes" (Netflix, 2021b).
Compared to Netflix's apparently inclusive and limitless offer, there is a host of much smaller, independent film streaming services that pride themselves on their exclusivity, their highly niche offer, their commitment to quality, and their reliance on human selection rather than software computation. Mubi, for instance, describes itself as "a curated online cinema where you can discover the world's greatest films" (Mubi, 2021a), a "hand-picked line-up" of titles "from everywhere on earth [...] Always chosen by us" (Mubi, 2021b). And, "if you are [...] hungry for an alternative to the mega-platforms, you'll find the carefully-curated, manageable selection [...] offered by YourScreen irresistible" (YourScreen, 2021). In a talk by YourScreen founder (Baker, 2021), entitled 'The Antidote to Netflix', the distinctiveness of the service is described in terms of "restricting choice to an exclusive handful of first-rate films" (my emphasis).

CURATING AN ANTIDOTE TO NETFLIX'S DISRUPTIONS
Curation has become a byword for resisting the power of massive, faceless, global entertainment corporations, and for promoting cultural value rather than shareholder profit (see e.g. Bhaskar, 2016). But in the era of #BlackLivesMatter and #MeToo, the indie-streamer focus on exclusivity rather than inclusivity raises some interesting questions about diversity, and about the nature of the personalised, on-demand culture. What is at stake, then, in the idea that the small, curated, niche streamers provide an antidote to businesses like Netflix? Antidotes are traditionally designed to cope with the effects of poison or infection, suggesting a very familiar trope in the centuries-long debate about the perceived dangers of new media forms that disrupt more established cultural activities. For some commentators, Netflix promotes an "addictive" and "bland" version of mass culture (Campbole, 2019). Even if there is some evidence of quality, Netflix is accused of discouraging discernment: "With so much quality content at our fingertips, will we be truly capable of appreciating the worth of what we are watching?" (Herberg, 2017). The idea that consuming certain forms of culture is bad for you is perfectly expressed in the moral panic about the dangers of Netflix-inspired binge-viewing: "Binge Watching Is Like a Drug" and can "be harmful to your health" (Antinozzi, 2021).
Implicit in arguments about the threats posed by Netflix, and the suggestion that an antidote is required are all sorts of assumptions about cultural value and taste, and the idea that Netflix is a marauding industrial behemoth. Netflix has also of course come under fire from those in the entertainment business. The corporation has certainly been disruptive in that context, harnessing new technologies to fire a broadside at the modern Hollywood studio-distributor oligopoly, leading to its characterisation as "the monster that's eating Hollywood" (Flint and Ramachandran, 2017). It has helped smash the distribution system of release windows, designed to protect the theatrical business by requiring a theatrical opening for all films, before they are allowed to show on other platforms. For some in the industry, then, Netflix is the enemy that destroyed a well-established, effectively functioning business model and a settled and lucrative relationship with consumers world-wide.
It has also disrupted the traditional idea of cinema as a shared, public, big-screen experience, for which some regard it as the enemy of film culture (e.g. Cosgrove, 2018). But of course, there is nothing new about watching big-screen films on small screens at home, and Netflix is simply one in a long line of disruptors going back to the rise of television in the 1950s, and the subsequent development of the home video market, then DVD and Blu-Ray. If cinema is the watching of films, Netflix has not destroyed cinema, it has simply consolidated the online relocation of the film experience.
As one of the foremost examples of "internet-distributed television" (Lotz, 2017), Netflix's global distribution and personalised, on-demand streaming have also disrupted the traditional national experience of broadcast television, with its linear arrangement of content in daily schedules. Again however, this is a process that can be traced back to the 1950s, when the US industry began exporting TV shows, and the late 1970s, when video began to make time-shifting possible. There is then nothing new about stopping and starting the flow of audio-visual content, watching such content in your own time, choosing what you watch and when. What is new about Netflix is the means of distribution, via the internet, the size of the catalogue from which one can choose content, and the development of a highly sophisticated algorithmic recommender system. To that extent, as Lotz (2017Lotz ( , 2020, Johnson (2019) and Lobato (2019) make clear, Netflix is a symptom of digitally-enabled media convergence and the wider development of on-demand culture.
That in itself does not mean that Netflix necessarily requires an antidote. Indeed, the numerous small, independent, niche platforms that promote themselves as discerning cultural antidotes to Netflix are themselves digitally enabled, on-demand providers. Their brand distinctiveness lies in how they rely on personal curation by passionate experts and offer films that are too specialised even for Netflix. But Netflix's algorithmic recommender system is itself just another form of curation, and it too offers certain films under the labels of 'arthouse', 'critically acclaimed' or 'award-winning', all markers of quality, taste and difference from the mainstream. It also presents content from all over the world and makes content available in a variety of languages.

GLOCALISATION AND DIVERSIFICATION
As a global corporation, Netflix is at the heart of the geopolitical debates about global media flows. For some commentators, this makes Netflix an arch example of media globalisation, "the perfect representation of American cultural imperialism" (Tardieu, quoted in Erbland, 2017). The argument is that its economic power contributes to the global domination of particular types of content and particular approaches to representation. For Netflix itself, however, the argument is that, precisely because it is a global player, it must acknowledge cultural distinctiveness and local specificity: "We understand [...] that there are cultural differences and some variances in content tastes around the world" (Netflix, 2021b).
Indeed, it is heavily invested in the business of diversification. Thus it offers a huge variety of films and series from which to choose. It supports 'local' production around the world, operating a textbook version of glocalisation, situating itself as both a global player and a local player. It supports diverse, innovative and challenging subject matter that would not otherwise have found its way to the screen. And it supports diversity and inclusion schemes, including its own Creative Equity Fund (Netflix, 2021e). In that respect, Netflix can be characterised much more positively as a provider of diverse story experiences to global audiences and an enabler of cultural diversity, not an enemy of culture.
For Barker and Wiatrowski (2017, p. 1): The brilliance of Netflix's strategy is in how its streaming video library manages to appeal to disparate groups of people across the world without a unified cache of content. Indeed the company takes the opposite approach, using its sophisticated algorithm and seemingly endless resources to buy, develop and distribute as many different types of content to as many micro-targeted audience groups as possible.
At the same time, however, Netflix (2021b) also argues: [...] that great, high-quality storytelling has universal appeal that transcends borders. That's why we are increasingly licensing and producing content all across the globe and Netflix members everywhere in the world can increasingly enjoy the same movies and TV series at the same time, free of legacy business models and outdated restrictions.
In other words, there is a tension between localisation and diversification on the one hand, and on the other hand the distribution and promotion of stories with so-called universal appeal that Netflix helps become globally dominant. Even then, diversity is an option: "With our global distribution, Netflix is well positioned to bring engaging stories from many cultures to people all across the globe" (Netflix, 2021b). In other words, the media flow is not one way.
NETFLIX'S "CONGLOMERATED NICHE" STRATEGY Lotz (2017, p. ch1) argues that Netflix follows a "conglomerated niche" strategy. As other broadcast, cable and internet channels and platforms became carefully branded niche providers, the Netflix strategy involved the conglomeration of multiple audience segments, which "achieves the advantages of scale while servicing heterogeneous tastes." For Lotz (2017, p. ch1), this is not "a 'mass' strategy", since there is no requirement or expectation that all Netflix viewers will engage with and value the same material. Nor is it a strategy based on targeting demographics. Rather, as Lotz (2020) explains, it is about targeting algorithmically identified niche sensibilities and taste clusters.
I would argue, however, that Netflix in fact operates a mixed economy, some of which cannot really be described as niche. Indeed, the corporation itself states that, "there are several facets to our content strategy. First and foremost, we want to deliver a wide variety of best-in-class stories that our members love and watch in big numbers" (Netflix, 2021d). By seeking to offer the biggest, most high-profile titles in as many territories as it can, it does therefore operate a "mass" strategy, or a form of broadcasting, circulating what it sees as, or what its audience data tell them, are the most universal stories. Second, Netflix want "to create great, locally authentic stories in countries all around the world" (Netflix, 2021d). By buying or developing a rich variety of lower-profile, often localised content, it thus operates a form of mixed programming. While this second level already exercises a degree of diversification, it is only at a third level that the niche programming really comes into play. By using its algorithms to recommend very particular content to individuals, it operates a form of narrowcasting.
The end result for Netflix is that they are able to provide what they see as "a personal experience" (Netflix, 2021b), but they can only do this through the intricate customisation of an industrial-scale library. The conjoining of the personal and the industrial, and of the local and the global, may seem like contradictions in terms, but this is the logic that defines the Netflix operation, and diversification is by default central to that operation. This focus on choice and personal agency is of course part of a much broader trend, the development of an industrial logic founded on the idea of "individual customization, rather than mass standardization" (Manovich, 2001, p. 29, in Alexander, 2016. There is, however, a distinction to be made between Netflix's "conglomerated niche" strategy and what we might call the "single-niche" strategy of the small, independent streamers that give space to a differently diverse range of films: Mubi, Dogwoof, Film Europe, Curzon, BFI Player, Indiepix, and others. In this respect, the streaming business has mirrored the film theatre business, with its separation of specialised, niche or art-house cinema from the mainstream and the multiplexes. These two different types of streamer operate different content acquisition and delivery strategies, and different systems for recommending viewing experiences. Netflix depends on its algorithm and its analysis of big data, whereas the indie streamers rely on what they call hand-picked curation. But these are in the end simply different means of curating content, different ways of filtering, selecting and prioritising. As I argued elsewhere (Higson, 2017): Both of these recommender systems assume an established individual taste that can be maintained and expanded through recommendations; both assume they know what is good for us. They offer different ways of shaping social experience, cultural taste and consumer behaviour.

GATEKEEPING AND CULTURAL VALUE
These are, then, two different means of cultural gatekeeping, two different models of cultural value -and, as Frey (2021b) makes clear, two different business models. The small, single-niche streamers promote themselves as the homes of great cinema, with their exclusive collections of hand-picked films dependent on human intervention and the exercise of taste, sophistication and judgement by passionate experts. Their modus operandi is presented as culture-led, rather than data-led. As the guardians of great cinema, they operate something akin to a traditional public service broadcaster, involved in the cultural transmission of what Matthew Arnold (2001Arnold ( , p. 2 [first published 1867) called "the best which has been thought and said" (Frey, 2021a). Such an approach begs all the usual questions about such cultural gatekeeping. Whose taste is being presented, to whom, and why? How inclusive is this concept of taste? How is it valued in relation to other taste formations, and why? Indeed, one might argue that the endless promotion of "the best which has been thought and said" is the perfect example of a filter bubble. This is surely implicit in the business models of some of the single-niche streamers, designed as they are to appeal to very specific audiences: the university-educated, cosmopolitan, urban-dwelling middle-class.
These 'anti-algorithmic' platforms set out to challenge what some see as the machine-based construction of cultural taste: the datafication of audiences (Livingstone, 2018), the mathematization of taste (Alexander, 2016), and the regulation of identity (Arnold, 2016, p. 58). From another perspective, however, Netflix might be seen as the perfect antidote to the patrician model of the well-educated, tasteful critic telling us what we should be watching. From this perspective, the cultural value of Netflix, with its vast library and the ability for audiences to select what they want to watch, depends on concepts of access, choice and consumer empowerment. Such concepts have long been central to those who have argued for the democratisation of technology and digital production and consumption.
As Hesmondhalgh (2019) points out, however, while "there is no doubt that digitalization has offered abundance, convenience and mobility of information and entertainment", the hopeful optimism "invested in digitalization as a means of media democratization [has] generally not been realized". The gathering and analysis of vast amounts of personal data, for instance, is a double-edged sword and has generated much debate, both for and against, while there is much that is problematic about the ownership and control of the various global media services.

ATTENDING TO THE LOCAL AS A GLOBAL CONTENT STRATEGY
Once again, however, we need to acknowledge the fact that diversification and attention to the local are central to the Netflix business model: "We're programming for diverse and varied tastes and for an increasingly global audience, so we have to be more reflective of the audience we serve" (Cindy Holland, VP of Original Content, in Is the Arab world..., 2018). Indeed, as Lobato (2019, pp. 14-16) notes, the more global Netflix has become, the more differentiated and localised it has become. The bottom line is that diversification of audiences through micro-targeting and personalised viewing recommendations is good for business, since it expands the number of potential revenue streams.
The acknowledgement of cultural differences can be seen in Netflix's work as a producer. Drawing on its very substantial production budget (US$17 billion in 2021 [Low, 2021]), Netflix has now invested in the creation of local content with local stars in at least 30 markets (Lobato, 2019, p. 14). Many of these productions are made in languages other than American English, including French, Italian, Mexican, Arabic, Japanese, Hindi, Telugu, Tamil, British English, Korean, Turkish and Brazilian Portuguese. This attention to the local is typical of global content strategies (Lieb, 2018). As one commentator puts it, companies like Netflix need "to put in the work to win over regional audiences with original content that feels less like a distant Hollywood import, and more like something that reflects the audience's culture and lived experience, and is accessible in their mother tongue" (Skinner, 2020). Not surprisingly, Netflix's investment in locally produced content is growing, in part driven by government regulations, quotas and competition from local platforms (Thomson, 2020): "In 2015, about 75% of Netflix's original content was American; now the figure is half" (Charlemagne, 2021).
Indeed, "local originals were the most popular titles in many countries, including India, Korea, Japan, Turkey, Thailand, Sweden and the United Kingdom in 2019" (Zacks, 2020). Lobato (2019, pp. 126-130) explains that Netflix realised that if they were to run a successful operation in the Japanese market, they would have to carry a far higher proportion of local content than they were used to. Meanwhile, in Europe Netflix became the largest single commissioner of new European scripted content in 2020, with 72 new commissions, significantly more than any of the national public service broadcasters (Kaminska, 2021). This was no doubt in part due to pressure on Netflix and other global streamers from the European Union to invest in and provide space for local content (Trinidad García Leiva and Albornoz, 2020), but it is also clearly a product of Netflix's strategy of glocalisation. Even so, locally produced content still represents very much a minority of the content in Netflix catalogues in all but the US market. In many other markets, "less than 10% of catalogues are locally-originated" (Thomson, 2020).
Netflix also adapts existing content for different markets, widening access by dubbing material on its platform in at least 34 languages and subtitling in a few more (Charlemagne, 2021), with Lobato (2019, p. 120) concluding that "Netflix is now translating more content into more languages than almost any other media service has done or could do".
Lobato's (2019) detailed study of Netflix's global operations demonstrates the way in which Netflix has been obliged to treat each nation or territory as its own niche, requiring content to be localised in very particular ways. For Netflix to be globally successful, it has thus had to take seriously the appeal of what Straubhaar (2003, p. 85) calls cultural proximity: "the tendency to prefer media products from one's own culture or the most similar possible culture". To respond to this tendency, Netflix has had to make compelling content locally accessible by "understanding and negotiating cultural differences" (Lobato, 2019, p. 114). Lobato (2019, p. 114) makes it clear just how challenging Netflix has found this business of localisation, with national markets needing to be understood "as discrete containers characterized by diverse tastes, income levels, languages, genre preferences, willingness to pay, and other factors". As a consequence, Netflix is valued differently and has a different cultural status in each national market. At the same time, it would be misleading to reduce niche targeting to geographically defined forms of cultural proximity. As Lotz (2020) and Straubhaar (2007) argue, taste clusters and communities also form around proximate or familiar genres, themes and values.

ARAB-LANGUAGE NETFLIX ORIGINALS
According to Netflix's Chief Product Officer, "Netflix members around the world want authentic storytelling, they want a perspective from a passionate creator that's grounded in the local culture" (Robinson, 2019). The way this strategy plays out can be seen in the case of Netflix's attempts to win audiences in the Arabic-speaking territories of the Middle East and North Africa (MENA). As Netflix went global in 2016, it started to seek out opportunities for local storytelling, which in this case, as Ahmed Sharkawi, Netflix Director of Arabic and African Originals, explained, meant "stories made by Arabs for Arabs, instilling pride, showcasing Arabic-language shows and films to the world, and providing Arab talent and filmmakers with a platform to gain fans globally" (Netflix airs..., 2020). And as Haddad and Dhoest (2021, p. 261) demonstrate, "the arrival of Netflix to the Arab world in 2016 was a game-changer".
Netflix has gradually built up a library of Arabic-language productions, including a series of Arab classics; they have localised content through subtitling and dubbing; and since 2018 they have released a number of their own 'original' productions, produced in Arab countries, by Arab creators, with production companies based in Jordan, Egypt, Saudi Arabia, Tunisia, Kuwait, Syria and the UAE, amongst other countries. They have sought to produce across a range of genres and to address a range of regional audiences. Jinn, a supernatural teen thriller series based in Jordan, and Justice, a UAE series about a young female lawyer, were among the first releases, in 2019. Dollar was shot in Lebanon by Syrian director Samer Berkawi. Paranormal, the first original Egyptian Arabic-language series, was adapted from a best-selling Egyptian thriller novel series. Al Rawabi School for Girls, a female-centric teen comedy, had a female production team, again based in Jordan. Drama Queen is an Egyptian comedy drama series with Abla Fahita, featuring her already established satirical puppet. Finding Ola is a comedy drama with Egyptian-Tunisian actress and executive producer, Hend Sabry. Netflix has also established multi-title partnerships with two Saudi Arabian companies, Myrkott and Telfaz11 (Vivarelli, 2020;Lakhpatwala, 2020;Farrow, 2021;Schumaker, 2020).
The commitment to diversity and local storytelling takes two forms. First, there is a desire to enable Arab creators from diverse backgrounds "to tell their stories in the most powerful form", and to provide those creators with the necessary resources, including "time to perfect ideas, and technical and creative resources to deliver the best version of their work" (Sharkawi, in Netflix airs..., 2020). Second, there is the audience: "as a global business it is incredibly important for us to make sure our slate reflects the diverse cultures and experiences of our [audience] members", and provides them with opportunities "to see their lives represented on screen" (Sharkawi, in Netflix airs..., 2020). Another notable feature of the Arabic-language Netflix originals is the commitment to female-led productions and female-centred stories: "empowering female creators is crucial" if we are "to give more people a chance to see their lives represented on screen" (Nuha El Tayeb, Director of Content Acquisitions for Netflix MENA, in Mzoughi, 2021).
Diversity has clearly been not merely an ethical commitment, but also very much a business strategy. As the leading US-based trade papers noted, "With over 400 million potential viewers, the [MENA] region has vast potential" (Gandin Mark, 2018), with those viewers already spending "an average of 19 hours per week watching scripted Arabic-language series" (Gandin Mark, 2018), and "SVOD [subscription video on demand] penetration [...] estimated to double to more than 27 million subscribers by 2025, with Netflix projected to be the MENA market leader" (Vivarelli, 2020).
Television production in Arab countries has for many years been dominated by the production of musalsalat, month-long series screened during Ramadan, catering to middle-aged female audiences above all. "Advertisers pour their biggest budgets into 30-day serialized dramas and comedies" (Gandin Mark, 2018). These dramas tend to be "heavy on plot and light on character development", but are rarely given large budgets, with 30-episode series often having just "a two-month turnaround, a process that would take more than a year in the American market" (Gandin Mark, 2018). By contrast, several of Netflix's Arabic-language originals were clearly pitched at younger audiences, "the region's fastest growing demographic segment. It is a generation that thirsts for authentic stories that mirror the diversity of their heritage and culture" (The story..., 2018).
The Netflix strategy of producing much shorter, more intense drama series is very different to the Ramadan TV model. In this case, the commitment to more complex, tightly-scripted and character-driven stories means "creators will no longer feel bound to unnecessarily stretch or abridge storylines" and will have time for collaboration, revision and innovation (Bedirian, 2020;Gandin Mark, 2018). The same applies to the unprecedented amount of visual and special effects in Paranormal: [...] the scale definitely allows the industry to dream bigger and bring in higher quality content as a series [...] We want to empower creators and give them the freedom to structure their vision however they see fit (Sharkawi, in Bedirian, 2020).
Feedback from the Arab creatives behind the various local Netflix productions makes clear that the appeal for them is the opportunity to work with decent-sized budgets, a commitment to finely tuned storytelling and high-quality production values. Mir-Jean Bou Chaaya, director and executive producer of Jinn, for instance, commented on the "great experience [of] working with Netflix" and the fact that it was "a very collaborative partnership where we have felt empowered to bring 'Jinn' to life through our own creative vision with authenticity at the heart of the show" (The story..., 2018). Elan Dassani, head writer and executive producer of the series, similarly commented that, "The freedom Netflix gives producers really allows us to delve deep into the nature of storytelling and bring Arabic stories to life" (The story..., 2018). Mohamed Hefzy, producer of Paranormal, noted that it was "a great learning experience on how to produce a high-quality, bigger budget premium series and see it launch in multiple territories simultaneously" (Netflix airs..., 2020).
Clearly, however, there is again a hard-edged business logic at work as well, with Holland explaining that commissioning is in part driven by their data about audience preferences: "We are able to make informed decisions by using our data to size the potential audience for a project and estimate the right level of investment" (Is the Arab world..., 2018). Evidently, the production of Jinn as a contemporary supernatural teenage drama in part built on the success of Stranger Things, the second series of which "became the region's [...] fastest binge-raced show" (How Netflix..., 2018).
Inevitably, some commentators and audiences in the Arab world were less positive about the productions in which Netflix invested. According to several sources, the Jordanian production Jinn was "greeted with near-unanimous negative reviews and widespread censure from public institutions and Islamic groups in the kingdom" (Fahim, 2019). Bou Chaaya, the director of the series, may have set out "to make a show that represents teenagers from the region [...] Not something with stereotypes or false representations written all over it" (Hawksley, 2019). One report, however, quoted tweets from viewers who felt the producers had "made no real effort to explore the country's traditions and culture, instead falling back on tropes familiar from US TV productions" (Robinson, 2019). Others felt that Netflix "is still an American-centric company, whose pitches and projects from local staff suffer US oversight, from character developments to plot details" (Hawksley, 2019). "Nothing feels authentic, emotionally real or believable" (Fahim, 2019). Even so, as Haddad and Dhoest (2021, p. 258) confirm, "Netflix has brought a new model to the Arab world that paved the way for Arab writers who are eager -even if only modestly -to change the status quo." Part of the problem for Netflix, however, is that they want to win over local audiences but they also want to create stories that will travel to other regions. As El Tayeb explains, "We want to empower Arab storytellers with the tools they need to tell the best version of their stories, and then help connect these stories with a global audience so we're able to export cultures that were previously underserved and underrepresented" (Lakhpatwala, 2020). In other words, narrowcast content needs also to be "universal" enough for it to be successfully broadcast.

INVESTING IN DIVERSITY
Alongside the production of so-called local content, Netflix has also taken risks by investing in innovative, difficult and challenging stories that arguably would not otherwise have found their way to the screen. Well-known examples include investment in Alfonso Cuaron's Roma and Martin Scorsese's The Irishman, while Black Mirror: Bandersnatch was an innovative interactive narrative. The development of its 'Strong Black Lead' campaign, promoting a wide range of Black narratives, is another good example. Miya Treadwell (2021) argues that, in this respect, "Netflix stood apart from competitors regarding its diversity and social engagement with a multitude of Black content creators." As she goes on to note, "the platform specifically highlighted its collaborations with director Ava DuVernay, whose work involves social issues underscoring Black life", with Netflix funding both 13 th and When They See Us.
In Nigeria, Añulika Agina (2021) explains that: [...] filmmakers are optimistic about Netflix because it offers deals and funding that allow their films to reach global audiences without sacrificing local audiences' access to the same films, and their profit.
Such filmmakers are not required to shy away from challenging subject-matter either: "Netflix's involvement affords opportunities to address controversial or vexed subjects that might not otherwise be told" (Agina, 2021). Ted Sarandos (2021), co-CEO of Netflix, explains that since their move into production, the company has "aspired to tell stories that weren't available elsewhere [...] in terms of the variety we offered [...] the people and cultures we brought to the screen" and the hiring of "talent from traditionally underrepresented communities". As an example of hiring underrepresented talent, he cites actor Laverne Cox, cinematographer Rachel Morrison, producer Yance Ford and screenwriter Dee Rees.
In 2021, Netflix and the USC Annenberg Inclusion Institute published a rigorous academic report commissioned by Netflix to look at the social inclusivity and diversity of Netflix productions, both on screen and behind the camera (Smith et al., 2021). This closely followed another Netflix report on what the company is doing to diversify its workforce and create a more inclusive workplace culture (Myers, 2021). Both of these are admirable initiatives, particularly since Netflix does not always appear in a brilliant light in the reports. In terms of their productions, they are doing comparatively better than some of their competitors in terms of gender equity and Black representation, but they also fall short in some areas, with many identities remaining largely invisible on-screen. In terms of their own workforce, the company has made efforts to become more inclusive, with staff noting that "the working culture at Netflix allows for diversity of thought, culture, background and perspective in its essence" (Omer, 2017). Even so, the party line is that, while "we've made good progress over the last three years [...] let's be clear, we're not where we want to be and we need to do better" (Myers, 2021). Stacy Smith, who led the USC study, commented that she had been: [...] encouraged by Netflix's efforts at being transparent with its data and putting that information directly in the hands of executives who are responsible for greenlighting projects and hiring talent. That's a good first step in holding themselves accountable (in Fang, 2021).
Netflix has also invested in a number of diversity and inclusion initiatives directed at the production community, most recently launching the Netflix Fund for Creative Equity, which will invest £100m over 5 years "in the next generation of underrepresented storytellers" (Sarandos, 2021). Some of those funds have already been committed to initiatives to support women filmmakers in France, Germany and Canada (Easton, 2021), to documentary filmmakers in the UK (Dalton, 2021), to emerging talent in sub-Saharan Africa (Obioha, 2021), and to indigenous content creators in Australia (Groves, 2021). The previous year, Netflix had committed another $100m to support creative workers impacted by the COVID-19 pandemic, with initiatives across the world, including Poland (Advanced Television, 2020), Nigeria (Dalberg, 2020) and Egypt (Egypt Today, 2021).

THE VALUE OF DIVERSIFICATION AS A BUSINESS STRATEGY: SOME CONCLUSIONS
While it is easy to be cynical about diversification as a business strategy, and while many would argue that the Netflix catalogue is simply not diverse enough, the initiatives outlined here are important, especially in a global corporation. Are they adequately played out in their algorithmic approach to the prediction and promotion of taste? One of the constant criticisms of this approach is that it is designed to reproduce existing taste by offering audiences more of what they already like, rather than pushing them out of their comfort zones or broadening their horizons. The algorithmic approach, it is argued, creates filter bubbles and echo chambers. From this perspective, "although Netflix's brand identity centers on notions of user choice, its algorithms work to actively negate choice" (Arnold, 2016, p. 59).
As Kitchens et al. (2020) note, "echo chambers and filter bubbles are potent metaphors that encapsulate widespread public fear" about the way that people use digital media systems, and the ways those systems work. The argument is that the algorithms used by such systems create "a narrowing of information diversity". In the case of Netflix, for Alexander (2016, p. 89), this means it is "less likely you will encounter any 'great films' outside your comfort zone". The recommender system, it is argued, effectively filters out content with which the user is unfamiliar, on the assumption that it is unlikely to appeal to the user. This is just one aspect of Netflix's role as a gatekeeper and tastemaker, mediating between producers and audiences, encouraging some content and some audiences through the various cultural gates, but stopping others from entering. Netflix also selectively markets some of its products, thereby trying to shape people's choices and reconstruct their taste -albeit according to their big-data analysis rather than individual whim. The arguments about filter bubbles are often highly speculative, however, with little empirical audience/user research to back up the claims. The idea that Netflix's recommender system involves a loss of serendipity and chance encounters with "great films" certainly assumes a very passive form of user experience and audience engagement, and a very fixed idea of cultural value. This is part of a wider trend, noted by Livingstone, whereby: [...] to theorize recent and profound changes, media scholars are reasserting monolithic accounts of power that tend to downplay or exclude audiences and the significance of their lifeworlds.
The counterweight to this trend is rigorous research, both qualitative and quantitative, about audiences, users and reception. One such piece of research (Frey, 2021a), suggests that few SVOD users in the UK actually rely on recommender systems and use a variety of other information sources, in particular word-of-mouth, to inform their viewing choices.
Netflix is a complex and ever-changing cultural space, and its huge library and constantlyupdating user interface create for the consumer an impression of limitless abundance, endless choice, diversity of material and customized, personalised experiences. In reality, as Lobato (2019), Alexander (2016) and others demonstrate, there are clear limits to what is available, the catalogue is finite, access is circumscribed and content is editorialised in various ways. National and regional policies and the subscription rates that Netflix charge in different territories also limit accessibility. In India, for instance, "Netflix is unaffordable for all but the most affluent Indians" (Lobato, 2019, p. 126), while the EU has sought to regulate video on demand services and limit their activities in various ways (Trinidad García Leiva and Albornoz, 2020). With the establishment of Disney+, Disney content is no longer available on Netflix, while there is a host of festival successes from around the world that never find their way onto the platform. Clearly, some content is considered too niche for Netflix to acquire. It is also evident that the promise of a long tail of niche-interest material always being available (Anderson, 2006) has not been fulfilled in quite the way that its proponents expected.
The shift that Netflix embodies, "from a mass economy into a niche market of personalized services" (Alexander, 2016, p. 84) evidently has its limits, especially when that process of industrial customisation is coupled with an investment in diversification. As Sarandos (2021) explains: Over the years, we've seen that to drive real change, we need to approach our work with an "inclusion lens." That means asking more questions like: "whose voice is missing? Is this portrayal authentic? Who is excluded?" Clearly some are still excluded, and the questions raised about choice, accessibility, the construction of taste and the acknowledgement of difference are nothing if not complex and multi-faceted. Netflix evidently plays a relatively powerful role in regulating what sort of audio-visual content audiences around the world engage with, as well as when, where and how. It may in various ways require an antidote, but it is also heavily invested in the business of diversification, and its commitment to access and inclusion has to be taken seriously.

ACKNOwLEDGEMENTS
Thanks to Miya Treadwell, Duncan Petrie, Agata Lulkowska and Siuda Piotr for commenting on drafts of this paper.