Netflix

Netflix, Inc.
Type	Public
Traded as	Nasdaq: NFLX S&P 500 component
Industry	Streaming media Entertainment Film and television production
Founded	29 August 1997, Scotts Valley, California
Founder	Reed Hastings Marc Randolph
Headquarters	Los Gatos, California, United States
Key people	Ted Sarandos (Co-chief executive officer) Greg Peters (Co-chief executive officer) Spencer Neumann (Chief financial officer)
Products	Netflix streaming service Netflix Games Netflix House
Employees	Approximately 14,000
Website	netflix.com

Netflix, Inc. is an American streaming and entertainment company headquartered in Los Gatos, California. It operates the largest subscription video service in the world, with more than 300 million paid memberships across over 190 countries, and it produces and commissions a volume of film and television that exceeds that of any traditional studio.

The company was founded on 29 August 1997 by Reed Hastings and Marc Randolph as a service that rented DVDs by post. It introduced a flat-rate monthly subscription with no late fees in 1999, launched streaming in 2007, began commissioning original programming in 2013, expanded to almost every country in the world in January 2016, and discontinued the DVD business on 29 September 2023 after shipping more than five billion discs.

Netflix is generally credited with, and sometimes blamed for, the transformation of how filmed entertainment is made, distributed and watched. It established the binge release, in which an entire season is published at once; it normalised subtitled foreign-language drama for English-speaking audiences at scale; it removed the scheduled broadcast as the organising principle of television; and its willingness to spend very large sums on content forced every major studio and broadcaster into a streaming strategy, an industry-wide reorganisation that has cost its competitors tens of billions of dollars and that Netflix itself survived by being early.

The company has also attracted sustained criticism: for cancelling series after one or two seasons in a way that its own data-driven commissioning makes predictable, for the effect of the streaming model on residual payments to writers and actors, for its tax arrangements in several countries, for the environmental cost of video delivery, and for individual programming decisions that have drawn boycotts from both political directions.

History

Founding, 1997 to 1999

The founding story that Netflix told for many years was that Reed Hastings conceived the company after being charged US$40 in late fees by Blockbuster for a copy of Apollo 13. Marc Randolph has said that the story is a useful simplification rather than a lie, and that the real origin was a series of conversations during the commute the two shared between Santa Cruz and Sunnyvale, in which they worked through a long list of possible mail-order businesses before arriving at video.

Both had done this before. Hastings, a mathematician by training and a former Peace Corps teacher in Eswatini, had founded Pure Software, a debugging tools company, and sold it to Rational Software in 1997 for about US$700 million, leaving him with capital and with a strong aversion to what he regarded as the bureaucratic process-heavy culture Pure had developed as it grew. Randolph had been a direct marketing executive and had co-founded MicroWarehouse.

The enabling technology was the DVD, which had launched in the United States in March 1997. It held a feature film, it was light enough to post for the price of a first-class stamp, and it did not break in the way a videocassette would. The founders tested the proposition by posting a compact disc to themselves in a greetings card envelope; it arrived the next day intact, and they proceeded.

Netflix.com launched on 14 April 1998 from Scotts Valley, California, with 30 employees and 925 titles, essentially the entire DVD catalogue then in existence. The initial model was conventional pay-per-rental with late fees, and it did not work well. Customers used it to try DVDs they could not find locally, then stopped.

The change that made the company was announced in September 1999: a flat monthly subscription giving unlimited rentals, with a limited number of discs held at a time, no due dates, no late fees and no shipping charges. The red prepaid envelope became the company's most recognisable object. By the end of 2000 it had abandoned single-rental sales entirely.

In 2000 Hastings and Randolph flew to Dallas and proposed that Blockbuster buy Netflix for US$50 million and run it as the online arm of the rental chain. Blockbuster's chief executive John Antioco and his team declined, and by several accounts found the proposal amusing. Blockbuster filed for bankruptcy in 2010. The meeting has become one of the most frequently cited episodes in business history, though the more informative detail is that Blockbuster did eventually build a competitive online service in 2004 and abandoned it under pressure from franchisees whose stores it was undermining, which is a more interesting failure than simple blindness.

The DVD years, 2000 to 2007

Netflix filed for an initial public offering in 2002 and listed on Nasdaq on 23 May at US$15 a share. Randolph, who had been chief executive at founding, had handed the role to Hastings in 1999 and left the company in 2003.

The DVD business was an operations business, and Netflix built it with unusual care. It opened distribution centres progressively across the United States so that a growing proportion of subscribers were within overnight postal reach, reaching more than fifty centres and next-day delivery for the large majority of the country. It worked closely with the United States Postal Service on envelope design, on automated sorting and on the tolerances that would allow discs to survive machinery, and it designed its warehouses around human hand-sorting because the volumes and the return flows did not suit automation.

The catalogue was its other advantage. Because Netflix held one national inventory rather than stocking thousands of shops, it could carry titles that no physical store could justify, and a large fraction of its rentals came from outside the current releases. Chris Anderson used the company as the central example of the long tail in his 2004 article and subsequent book, and the argument that aggregate demand for obscure titles could exceed demand for hits became one of the defining ideas of early internet commerce.

Cinematch, the recommendation system, was the mechanism that made the long tail usable, since a catalogue of 100,000 titles is worthless if a subscriber cannot find anything in it. In October 2006 Netflix announced the Netflix Prize, offering US$1 million to any team that could improve the accuracy of its predictions by ten per cent. More than 40,000 teams from 186 countries entered. The prize was won in September 2009 by BellKor's Pragmatic Chaos, an ensemble of three teams, and the competition is generally credited with advancing collaborative filtering and matrix factorisation techniques substantially. Netflix did not deploy the winning algorithm, on the grounds that the engineering cost outweighed the benefit and that its business was by then moving to streaming, where the available signals were different. A planned second prize was cancelled after researchers demonstrated that the anonymised competition data could be de-anonymised by cross-referencing public film ratings, and a lawsuit followed.

Blockbuster launched Blockbuster Online in 2004 with a combined online and in-store offering that Netflix could not match, and a price war through 2005 and 2006 cost both companies. Blockbuster's Total Access programme was genuinely effective and was curtailed by internal conflict, by the debt the company carried after its separation from Viacom, and by the ousting of Antioco in 2007. Walmart also entered and exited DVD rental, handing its subscribers to Netflix in 2005.

By 2007 Netflix had shipped its billionth disc and had around 6.3 million subscribers.

The move to streaming, 2007 to 2012

Netflix launched its streaming service, then called Watch Instantly, in January 2007. It was included free with a DVD subscription, offered about 1,000 titles against the 70,000 available on disc, ran only in a browser on Windows using a proprietary plug-in, and gave subscribers a monthly hour allowance proportional to their plan.

Hastings had anticipated the transition from the beginning, which is why the company was named Netflix rather than something involving discs, and had told investors for years that mailing physical objects was an interim arrangement. The internal discipline that mattered was the willingness to damage the profitable business in order to build the one that would replace it.

The streaming service expanded quickly onto devices. Netflix partnered with LG and then with Roku, which had begun as an internal Netflix hardware project and was spun out in 2008 so that Netflix would not become a hardware company or antagonise other manufacturers. The Xbox 360 gained Netflix in 2008, the PlayStation 3 and Wii in 2010, and by 2011 the application was embedded in most televisions and Blu-ray players sold in North America. In 2010 Netflix separated its streaming plans from its DVD plans for the first time and launched in Canada, its first international market, followed by Latin America and the Caribbean in 2011 and the United Kingdom, Ireland and the Nordic countries in 2012.

Qwikster

In July 2011 Netflix announced that it would separate DVD and streaming into two subscriptions, which for a subscriber taking both raised the price from US$9.99 to US$15.98, an increase of sixty per cent. In September, Hastings announced that the DVD business would be spun into a separate company and website called Qwikster, with a separate queue, separate billing and separate ratings.

The reaction was severe. Netflix lost approximately 800,000 subscribers in the third quarter, and the share price fell by more than three quarters over the following months, from around US$300 to under US$70. Hastings issued an apology in which the tone, which began by explaining that he had messed up and proceeded to restate the reasoning that had caused the problem, was widely regarded as making matters worse. Qwikster was cancelled on 10 October 2011, three weeks after it was announced, and never launched. The price separation remained.

Hastings has discussed the episode repeatedly since and has drawn two conclusions from it. The first is that the strategic reasoning was correct, in that the two businesses had different economics and needed different management, and that the company's later success depended on the separation. The second is that he had become so convinced of the strategic case that he did not test how it would be received, and that a company whose culture selects for conviction needs a deliberate mechanism for dissent. The Qwikster memo had circulated internally and almost nobody had objected, which he has described as a failure of the culture rather than of any individual.

The recovery was rapid in business terms and slow in reputational ones. The share price did not regain its 2011 peak until 2013, by which point the reason had changed entirely.

Original programming, 2013 to 2016

Netflix's move into commissioning was forced rather than chosen. As streaming grew, the studios that licensed content to it recognised that they were supplying a competitor and raised prices or withdrew. Starz ended its agreement in 2012, removing a large part of the film catalogue. The only durable answer was to own the content.

The first commission of consequence was House of Cards, for which Netflix outbid HBO and AMC in 2011 with a commitment to two seasons and roughly US$100 million without seeing a pilot. The company's stated reasoning, which became one of the most repeated claims in the industry, was that its data showed an overlap between subscribers who watched David Fincher films, subscribers who watched the original British series and subscribers who watched Kevin Spacey. The claim was true in outline and has been overstated in retelling; the more accurate description is that the data reduced the perceived risk of a decision that was still a judgement.

The whole first season was released on 1 February 2013. The decision to publish all thirteen episodes at once was a deliberate break with a century of scheduling practice, and its justification was that Netflix had no advertising to sell and therefore no reason to ration attention, and that its subscribers had demonstrably been watching whole seasons of licensed shows in single sittings. Binge-watching entered general usage within the year.

Orange Is the New Black followed in July 2013 and became the most-watched original of the platform's early years. Arrested Development was revived in May 2013 after Fox had cancelled it in 2006, establishing a pattern of resurrecting cancelled shows with existing audiences. Narcos, Master of None, Stranger Things in 2016, The Crown in 2016 and BoJack Horseman followed.

In September 2013 House of Cards won three Primetime Emmy Awards, the first major Emmys for a programme distributed online, and the Television Academy's willingness to treat streaming as television was settled.

Film came more slowly and more contentiously. Beasts of No Nation in 2015 was released simultaneously in cinemas and on the service, which caused the major American theatre chains to refuse it. The exclusivity window, under which a film plays in cinemas for around ninety days before any home release, was the mechanism Netflix was attacking, and the argument has continued since. The Cannes Film Festival changed its rules in 2018 to require French theatrical release for competition entries, which removed Netflix from competition. The Academy of Motion Picture Arts and Sciences debated similar restrictions after Roma, directed by Alfonso Cuarón, won three Oscars in 2019, and Steven Spielberg argued publicly for them.

The international expansion was the other decisive move of the period. Netflix launched in Japan in September 2015 and then, at the Consumer Electronics Show on 6 January 2016, Hastings announced that the service had gone live in 130 additional countries simultaneously, bringing it to more than 190. China, Crimea, North Korea and Syria were the exclusions. The engineering required to launch everywhere at once, including payments, subtitling, content rights and delivery, had taken years to prepare.

Global production and the co-chief executive era, 2017 to 2020

From 2017 Netflix's strategy shifted from exporting American programming to commissioning locally in each major market, on the reasoning that a service costing roughly the same everywhere must be worth the price everywhere, and that imported content would not achieve that.

The results included Dark from Germany, Money Heist from Spain, whose global audience far exceeded its domestic one and which was effectively rescued from a modest Antena 3 run, Sacred Games from India, Élite from Spain, Lupin from France and Sintonia from Brazil. By 2020 Netflix was one of the largest commissioners of scripted television in several countries where it had existed for only four years, a position that provoked both enthusiasm from producers and concern from regulators about cultural sovereignty. The European Union imposed a quota in 2018 requiring that at least thirty per cent of a streaming catalogue in Europe be European works.

Ted Sarandos, who had joined in 2000 from a video distribution background and had built the content organisation, was appointed co-chief executive alongside Hastings in July 2020, a structure the company said reflected how it had actually been run for years.

Spending rose steeply. Content expenditure passed US$12 billion in 2018 and approached US$17 billion by 2021, funded substantially by high-yield debt, because the accounting for content investment meant that a company growing rapidly would show negative free cash flow for as long as it was expanding. Analysts divided sharply over whether this was a sound investment in a durable library or a subsidy that would have to end, and the question was not settled until the cash flow turned positive in 2022.

Competition arrived in force in the same period. Disney+ launched in November 2019 and reached 100 million subscribers faster than Netflix had; Apple TV+ launched the same month; HBO Max and Peacock in 2020; Paramount+ in 2021. Disney's withdrawal of its library from Netflix, and NBCUniversal's withdrawal of The Office and WarnerMedia's of Friends, removed the most-watched licensed titles on the service, both of which Netflix paid large sums to retain temporarily while its own catalogue matured.

The COVID-19 pandemic produced a surge of subscriptions in the first half of 2020, adding more than 25 million in six months, and a production shutdown whose effects appeared in the release schedule about a year later.

The correction, 2021 to 2022

Squid Game, a South Korean series released in September 2021, became the most-watched title in the company's history, reaching approximately 142 million households in its first four weeks. It had been rejected by Korean broadcasters for a decade, was made for about US$21 million, and Netflix internally estimated its value at close to US$900 million. It settled the argument about whether English-speaking audiences would watch subtitled drama, and it also drew attention to the terms on which Netflix commissioned, since the creator Hwang Dong-hyuk received no additional payment for the success under the flat-fee production model then standard.

In November 2021 Netflix launched Netflix Games, included with a subscription and free of advertising or in-app purchase, beginning on Android and extending to iOS and later to televisions. It acquired the studios Night School, Next Games, Boss Fight Entertainment and Spry Fox, and hired Mike Verdu from Facebook to run it. Progress has been slow and the strategy has been revised repeatedly.

Then, in April 2022, Netflix reported that it had lost 200,000 subscribers in the first quarter, its first decline in more than a decade, and forecast the loss of two million more. The share price fell by about 35 per cent in one day and by roughly three quarters over the year. The company attributed the loss to market saturation, competition, the suspension of its Russian service after the invasion of Ukraine, and password sharing, which it estimated affected more than 100 million households.

The response reversed two positions the company had held publicly for years.

Advertising. Hastings had said repeatedly and emphatically that Netflix would never carry advertising, on the grounds that it corrupted the relationship with the viewer and that the company's simplicity was part of its value. In 2022 Netflix announced an advertising-supported tier, chose Microsoft as its technology partner over Google and Comcast, and launched Basic with Ads in twelve markets in November 2022 at US$6.99 in the United States. Hastings said he had been wrong. By May 2026 the advertising tier had approximately 250 million monthly active viewers, and Netflix had replaced the Microsoft arrangement with its own advertising technology.

Password sharing. Netflix had tweeted in 2017 that "love is sharing a password" and had tolerated the practice as marketing. From 2023 it began charging for members outside a household, first in Latin America and then globally. The measure was widely predicted to cause cancellations; instead it produced the largest subscriber additions in the company's history, as a substantial proportion of borrowers converted rather than leaving.

The company also conducted layoffs in 2022, roughly 450 people across two rounds, the first significant reductions in its history, and shut down several units including the original editorial site Tudum in its first incarnation.

Scale and live events, 2023 to 2026

Hastings stepped down as co-chief executive in January 2023 to become executive chairman, and Greg Peters, formerly chief operating and product officer, joined Ted Sarandos as co-chief executive. Eunice Kim became chief product officer and Elizabeth Stone chief technology officer in October 2023.

In April 2023 Netflix announced that the DVD service would end. The last discs were sent on 29 September 2023, with subscribers told they could keep what they held. Over 25 years the service had shipped more than five billion discs.

The strategic direction since has been toward live programming, events and sport, which Netflix had previously avoided on the grounds that live content is expensive, perishable and technically demanding.

In January 2024 Netflix agreed a deal with WWE valued at a reported US$500 million a year over ten years for the weekly programme Raw from 6 January 2025, initially in the United States, Canada, the United Kingdom and Latin America, together with international rights to SmackDown, NXT and pay-per-view events and the former WWE Network library. It later took the library rights in the United States as well, displacing Peacock.

On 15 November 2024 Netflix streamed a boxing event from AT&T Stadium in Arlington, Texas, headlined by an exhibition between Jake Paul and Mike Tyson and by Katie Taylor against Amanda Serrano for the undisputed lightweight titles. The stream suffered significant technical problems, with buffering and failures reported widely, but peak concurrent viewership was reported at 65 million, surpassing the 2023 Cricket World Cup final on Disney+ Hotstar as the most live-streamed sporting event. Netflix reported an average minute audience of 108 million worldwide and said the Taylor versus Serrano bout, with 47 million in the United States, was the most-watched women's professional sporting event in American history.

On 25 December 2024 Netflix carried its first NFL games, the Kansas City Chiefs against the Pittsburgh Steelers and the Baltimore Ravens against the Houston Texans, each averaging more than 30 million global viewers, and holds rights to at least one Christmas game in 2025 and 2026. In December 2024 FIFA announced that Netflix would be the exclusive American broadcaster of the 2027 and 2031 Women's World Cups. In August 2025 it took exclusive Japanese rights to the 2026 World Baseball Classic, all 47 games, its first live event in Japan. In November 2025 it acquired selected Major League Baseball events for 2026 to 2028, including an Opening Night game, the Home Run Derby and the Field of Dreams game.

Netflix reported in January 2025 that it had passed 300 million paid memberships, having added a record 18.9 million in the final quarter of 2024 and 41 million across the year. It then announced that it would stop reporting subscriber numbers quarterly, directing attention to revenue and margin instead, a change critics read as removing the metric by which the company had always been judged.

Other initiatives have extended the business beyond the screen. Netflix House, announced in June 2024, places permanent entertainment venues with retail, restaurants and interactive experiences at the King of Prussia Mall in Pennsylvania and Galleria Dallas, with a Las Vegas Strip site planned for 2027. The company produced Patriots on Broadway with Peter Morgan in February 2024 and Stranger Things: The First Shadow in London. It runs the Netflix Is a Joke comedy festival, from which The Roast of Tom Brady and Katt Williams's Woke Folk streamed live in May 2024. From late 2025 it moved into video podcasts, licensing shows from iHeartMedia, Barstool Sports and The Ringer and then commissioning its own.

In November 2024 Netflix discontinued its interactive specials, removing all but four and citing a wish to concentrate technical effort elsewhere. In May 2025 it redesigned the home screen for the first time since 2013, and in April 2026 introduced a vertical short-video feed called Clips on mobile.

The largest corporate event of the period was the attempt on Warner Bros. Discovery. On 5 December 2025 Netflix initially prevailed in bidding for WBD's studios and streaming businesses, including the Warner Bros. film, television and games studios, HBO and HBO Max and their libraries, and DC, at US$27.75 a share, about US$72 billion, valuing the whole of WBD at US$82.7 billion. Paramount Skydance then made hostile all-cash counter-offers for the entire company, arguing fewer regulatory obstacles. On 26 February 2026 WBD agreed to be acquired by Paramount Skydance at US$31 a share, valuing it at US$111 billion. Netflix withdrew, saying its own offer was no longer financially attractive, and received a US$2.8 billion break fee.

Reed Hastings announced in April 2026 that he would leave the board in June. In May 2026 the company published The Netflix Effect, claiming a contribution of US$325 billion to the global economy since its 2016 international expansion, US$135 billion invested in film and television, and 425,000 jobs created. In July 2026 the Wall Street Journal reported that Netflix was examining live linear channels and third-party streaming bundles to address declining engagement and to strengthen advertising, and that it was among parties including Sony and Paramount in early discussions to buy the film discovery platform Letterboxd.

Content

Commissioning and the use of data

Netflix's approach to deciding what to make is the aspect of the company most discussed and least accurately described.

The service records what is watched, when, for how long, where viewing stops, what is abandoned, what is rewatched and which images a subscriber clicked to start. This produces a picture of demand far more detailed than the ratings panels on which broadcast television relied. What it does not produce is a picture of demand for something that does not exist, and Netflix executives have said consistently that data informs the size of a bet rather than whether a creative idea is good.

What data genuinely changed is the structure of the commission. A broadcaster orders a pilot, tests it, orders a partial season, and cancels mid-run if the ratings fail, because the advertising it sells depends on the audience each week. Netflix sells no advertising against its original programming in the way a broadcaster does and has no schedule to protect, so it orders complete seasons, skips pilots, and has on several occasions ordered two seasons at once. This is more attractive to creators and is one of the principal reasons Netflix won projects against better-established buyers in its first decade.

The corresponding cost is the cancellation pattern. Netflix evaluates a series on the number of subscribers who watched it relative to what it cost, and applies that test after one or two seasons. A show with a devoted but modest audience, which a broadcaster might have kept for its stability, fails the test. Sense8, The OA, Mindhunter, 1899, GLOW, Santa Clarita Diet and a long list of others were ended abruptly, in several cases on a cliffhanger, and the resulting reluctance of some viewers to begin a Netflix series at all has been described by the company itself as a real problem.

The company's measurement has changed repeatedly, which has complicated comparison. It originally counted a view as two minutes, then moved to total hours viewed, then to hours divided by runtime, and since December 2023 has published What We Watched, a twice-yearly engagement report covering every title watched more than 50,000 hours. The first report, for the first half of 2023, found The Night Agent the most-watched title globally. The disclosure was a substantial change for a company that had for years refused to publish any audience figure, and was driven in part by pressure from the Hollywood unions during the 2023 strikes.

Programming

Netflix commissions across essentially every category. Its drama includes Stranger Things, The Crown, Bridgerton, Wednesday, Ozark, Squid Game, Money Heist and The Witcher. Its comedy and unscripted output ranges from Love Is Blind and Too Hot to Handle to Nailed It!, Drive to Survive, which is credited with a substantial increase in American interest in Formula One, and a very large volume of stand-up specials, on which the company spent aggressively and which made it the dominant buyer in that form.

Documentary has been a particular strength, with Making a Murderer, Tiger King, Our Planet, 13th and My Octopus Teacher, the last winning the Academy Award in 2021.

Animation has been rebuilt more than once. After an initial expansion, Netflix restructured its animation division in October 2023 and signed a multi-film agreement with Skydance Animation beginning with Spellbound (2024), partly replacing an arrangement Skydance had held with Apple. KPop Demon Hunters (2025) became the most popular film in the company's history, an outcome that had not been predicted internally and that produced a soundtrack which topped music charts, illustrating both the reach of the platform and the limits of its forecasting.

Film has included Roma, The Irishman, Marriage Story, The Power of the Dog, All Quiet on the Western Front and Glass Onion, alongside a large volume of action films made primarily for volume rather than acclaim, of which the Red Notice and Extraction franchises are examples.

Output agreements and overall deals have bound a number of individuals to the company, including Shonda Rhimes, Ryan Murphy, Barack and Michelle Obama's Higher Ground, David Benioff and D. B. Weiss, and Adam Sandler's Happy Madison, whose films have been among the most-watched on the service while receiving almost uniformly poor reviews, a divergence the company has cited as evidence that critical reception and subscriber value are different measurements.

Live and sport

Netflix resisted live sport for most of its history on the grounds that rights are expensive, expire, and confer no library value. Its entry has been selective rather than comprehensive: individual events with global attention rather than season-long league rights. The WWE arrangement is the exception and is structurally closer to a weekly entertainment series than to a sports rights deal.

The technical demands are considerable. Netflix's delivery architecture was designed for on-demand video, where content can be placed close to viewers in advance and where a delayed start is invisible. Live streaming inverts every assumption, and the Paul-Tyson event demonstrated the consequences publicly. The company has since invested heavily in live infrastructure and has delivered subsequent events including the NFL games without comparable failure.

Technology

Netflix accounts for a substantial share of downstream internet traffic in most countries in which it operates, and the engineering required to deliver it has been unusually influential because the company has published most of it.

Open Connect

Netflix does not buy delivery from a commercial content delivery network. It built Open Connect, its own network, and gives the hardware away.

An Open Connect Appliance is a server holding a large cache of Netflix's catalogue, installed free of charge inside an internet service provider's own network or at an internet exchange. When a subscriber presses play, the video is served from a machine inside their provider's infrastructure rather than traversing the public internet. The catalogue is updated overnight during off-peak hours, and the selection on each appliance is predicted from local viewing patterns.

The arrangement suits both sides. Netflix gets better performance and lower cost; the provider avoids carrying an enormous volume of traffic across expensive transit links. Thousands of appliances are deployed in more than a thousand networks. The scheme has also been a diplomatic instrument in disputes over network neutrality and interconnection fees, since a provider demanding payment for Netflix traffic is being offered a free way to eliminate most of it.

Encoding

Netflix encodes every title many times, at multiple resolutions and bitrates, and selects between them continuously during playback according to available bandwidth.

Its most significant technical contribution here is per-title and later per-shot encoding. Conventional practice applied a fixed bitrate ladder to all content, which wastes bandwidth on simple material such as animation and starves complex material such as a handheld night scene. Netflix instead analyses each title, and subsequently each shot, and assigns bitrates accordingly, reducing bandwidth by a substantial margin at equal quality. It also developed VMAF, a perceptual video quality metric that predicts how a human viewer will judge an encode, released as open source and now widely used across the industry.

The company adopted and promoted AV1, the royalty-free codec developed by the Alliance for Open Media, of which it is a founding member, and has pushed for its support in hardware.

Recommendations and presentation

The recommendation system determines almost everything a subscriber sees. The home screen is assembled individually: which rows appear, in what order, which titles populate each row, and in what order within the row.

Artwork is personalised as well. Netflix generates many images for each title and selects among them per viewer, so that a subscriber whose history suggests an interest in a particular performer may see artwork featuring them, while another sees an image emphasising the genre. The technique measurably increases the rate at which titles are started and has been criticised for occasionally misrepresenting a film, most often when a minor performer is given prominence to appeal to a particular audience.

The system has moved from the collaborative filtering of the Netflix Prize era to deep learning models trained on behavioural sequences, and its objective is long-run retention rather than immediate clicks, a distinction the company has written about at length.

Infrastructure and engineering culture

Netflix moved its infrastructure entirely to Amazon Web Services between 2008 and 2016, a migration prompted by a database corruption in August 2008 that halted DVD shipping for three days, and it was among the first large companies to commit to public cloud without retaining its own data centres, other than Open Connect.

Its engineering practices have had influence disproportionate to its size. Chaos Monkey, released in 2011, deliberately terminates production instances at random to ensure that systems tolerate failure, and the broader Simian Army extended this to whole availability zones; the practice is now known as chaos engineering and is widely adopted. The company published Hystrix for fault isolation, Zuul as an edge gateway, Eureka for service discovery, Spinnaker for deployment with Google, Conductor for orchestration, Atlas for telemetry, and Metaflow for machine learning workflows. Its microservice architecture and its writing about it were central to the popularisation of that pattern.

Netflix also publishes on media engineering through its technology blog, and its work on subtitle and dubbing pipelines, on quality control automation, and on production tooling has become a reference for the industry.

Artificial intelligence

Netflix has used machine learning throughout its operations for many years, in recommendation, encoding, artwork selection, demand forecasting and production scheduling.

Its use of generative techniques in content has been more contentious. The company confirmed in 2025 that generative tools had been used in production on a small number of titles, including for a visual effects sequence in the Argentine series El Eternauta, and published guidance for producers on permitted uses. It has stated that the technology is an efficiency in production rather than a replacement for writers or performers, a position the Writers Guild and SAG-AFTRA agreements from 2023 constrain in specific ways. The documentary What Jennifer Did drew criticism in 2024 over images that appeared to be generated, which the company did not adequately explain.

Business and finances

The model

Netflix sells a monthly subscription and, since 2022, advertising against a cheaper tier. It carries no commitment period, cancellation is immediate and unconditional, and this has been a deliberate feature since 1999: the company's position is that a subscriber who cannot easily leave is a subscriber who will resent staying.

The economics rest on a simple relationship. Content is a fixed cost, essentially independent of how many people watch, so every additional subscriber at a given level of spending is almost pure margin. A service with 300 million members can therefore spend on a single title what a national broadcaster spends in a year, and can spread that cost across an audience no national broadcaster can reach. This is the structural advantage that competitors with smaller subscriber bases have been unable to overcome, and it is why the streaming market consolidated rather than fragmenting further.

The accounting has confused observers for years. Netflix capitalises content spending and amortises it over the period it expects the title to be watched, which is front-loaded. During rapid expansion the cash paid out exceeds the amortisation recognised, so the company reports profits while generating negative free cash flow. Between 2013 and 2019 Netflix funded that gap with high-yield debt, accumulating more than US$15 billion, which sceptics read as evidence the model did not work and which management described as the cheapest available financing for a library of assets. Free cash flow turned durably positive in 2022 and the debt has since been reduced.

Plans and pricing

The service offers, in most markets, an advertising-supported tier, a standard tier and a premium tier differing in resolution and the number of simultaneous streams. Prices have risen repeatedly and Netflix has been open about its intention to raise them as the catalogue grows, a practice made possible by low churn and one that has periodically provoked public complaint without visibly affecting subscriptions.

The paid sharing feature, introduced globally in 2023, allows a member outside a household to be added for an additional fee, and the household is defined by network and device signals in a way that has caused friction for people who travel or who live across two homes.

Advertising has grown faster than expected. The tier reached approximately 250 million monthly active viewers by May 2026, making Netflix a significant seller of video advertising within four years of entering a business it had said it would never enter.

Results and reporting

Netflix passed 200 million members in January 2021 and 300 million in January 2025. Revenue has grown from about US$1.2 billion in 2009 to approximately US$39 billion in 2024 and higher since, with operating margin rising from single digits to the high twenties as the spending growth slowed relative to revenue.

The company stopped reporting quarterly subscriber figures from 2025, arguing that with multiple price tiers, advertising revenue and varying regional pricing, membership count no longer described the business. The decision removed the number by which investors and the press had measured Netflix for two decades and was criticised on that basis.

Netflix has never paid a dividend and has bought back shares substantially since 2021.

Corporate culture

Netflix's internal culture is unusually explicit and has been more widely copied, and more widely criticised, than that of any comparable company.

In 2009 the company published a 125-slide presentation titled Netflix Culture: Freedom and Responsibility, prepared largely by Hastings and the chief talent officer Patty McCord. It was viewed more than 20 million times and Sheryl Sandberg described it as possibly the most important document ever to come out of Silicon Valley.

Its principal assertions were these. The company wanted only high performers, described as a professional sports team rather than a family, and the appropriate response to adequate performance was a generous severance package. Managers should apply the "keeper test", asking whether they would fight to retain a given person and acting accordingly if the answer was no. Process should be minimised: Netflix has no formal holiday policy, no expense policy beyond "act in Netflix's best interests", and no travel approval. Compensation should be top of market in cash rather than in options, with employees free to choose the split. Context should be provided rather than control exercised, and decisions should be taken by the person closest to the information rather than escalated.

Two further practices distinguish it. Information is shared far more widely than is normal for a public company, including financial results before publication and details of commissioning decisions, on the argument that people cannot exercise judgement without context. And feedback is expected to be direct, frequent and unsoftened, including upward, with formal mechanisms including written 360-degree reviews visible to the subject and, in some parts of the company, live feedback sessions.

Hastings expanded the philosophy in No Rules Rules (2020), written with the business academic Erin Meyer, which added an account of how the model had been adapted, sometimes substantially, in the countries into which Netflix had expanded, since direct feedback is received very differently in Japan or the Netherlands than in California.

The criticisms are consistent. The keeper test produces anxiety that reduces exactly the risk-taking the culture claims to want. The absence of process disadvantages people who are less confident or less well networked, which in practice often means people who are not already like the existing staff. Termination is rapid and the stated commitment to candour can serve as cover for harshness. McCord herself left in 2012. Defenders answer that the arrangement is honest about a bargain most companies make implicitly, that compensation is correspondingly high, and that people who dislike it leave quickly.

Availability and access

Netflix is available in more than 190 countries. It has never operated in mainland China, where it attempted entry and settled for licensing individual titles to iQiyi, and it does not operate in North Korea, Syria or Crimea. It suspended its Russian service in March 2022 following the invasion of Ukraine.

Pricing varies widely by market, from a few dollars a month in India, where the company introduced a mobile-only plan in 2019 after finding its standard pricing far above what the market would bear, to considerably more in western Europe and North America.

The application runs on smart televisions, streaming devices, games consoles, set-top boxes, mobile phones, tablets and web browsers, and the company operates a device certification programme that has made Netflix support a standard requirement for television manufacturers. A dedicated Netflix button on remote controls, paid for by the company, has been a durable distribution advantage.

Downloads for offline viewing were introduced in 2016 after years of refusal, and smart downloads, which manage a queue of episodes automatically, followed.

Accessibility work includes audio description, closed captioning and dubbing at large scale; Netflix dubs and subtitles across dozens of languages and has been a substantial force in professionalising that industry, including publishing its own subtitling style guides and building a vendor certification programme.

Criticism

Effect on the industry

The most substantial criticism of Netflix is that its model has made filmed entertainment less sustainable as a career while making more of it than ever.

The traditional system paid writers, directors and performers residuals when a programme was repeated, syndicated or sold abroad, which converted a successful show into an income over many years. Streaming replaced that with a larger upfront fee and little or nothing afterwards, and because streamers did not publish viewing figures, there was no basis on which to negotiate a share of success. Seasons also shortened, from twenty-two episodes to eight or ten, reducing the number of weeks a writer is employed, and the practice of the "mini-room", in which a small group develops scripts before a series is greenlit, further reduced paid work.

These issues were central to the 2023 strikes by the Writers Guild of America and SAG-AFTRA, which together halted American production for months. The settlements introduced a streaming viewership bonus, minimum staffing in writers' rooms, and protections concerning artificial intelligence, and obliged the streamers to share viewership data with the guilds. Netflix, as the largest and most visible streamer, was the focus of much of the anger, and Sarandos's public remarks during the dispute were criticised as dismissive.

Cinema owners regard the company as an existential threat, and its refusal to accept the standard theatrical window has kept it out of most large cinema chains. The disappearance of the mid-budget adult drama from cinemas, and its reappearance on streaming, is a change Netflix accelerated and did not begin.

Cancellations and the catalogue

The pattern of cancelling series after one or two seasons has produced durable audience resentment and, in the company's own assessment, a measurable reluctance to start new Netflix shows. The corresponding practice of removing titles from the service when a licence expires, including in some cases the company's own originals removed for tax reasons, has also drawn criticism, since subscribers reasonably assume that a service that presents itself as a library is one.

Content controversies

Netflix's scale guarantees that some part of its catalogue offends someone, and it has faced organised campaigns from several directions.

Cuties (Mignonnes), a 2020 French film critical of the sexualisation of young girls, was promoted by Netflix with artwork that reproduced exactly what the film criticised. The company apologised for the marketing, kept the film, and faced a criminal indictment in Tyler County, Texas, later dismissed, along with a substantial wave of cancellations.

Dave Chappelle's special The Closer (2021) prompted protests by transgender employees and a staff walkout. Sarandos initially defended it in internal messages that argued content does not cause real-world harm, then said he had "screwed up" in how he communicated while maintaining the decision to carry the special. The episode led Netflix to add a line to its culture document stating that employees may have to work on content they find objectionable.

13 Reasons Why (2017) was criticised by mental health organisations for its depiction of suicide; Netflix removed the scene in question two years later after a study associated the release with an increase in adolescent suicides, a finding later contested.

The Crown has been criticised, including by former ministers and by Judi Dench, for presenting invention as history without a disclaimer; Netflix added one to later seasons.

Regional removals at government request, including an episode of Patriot Act with Hasan Minhaj in Saudi Arabia, have drawn accusations that the company's stated commitment to creative freedom is contingent on market access.

Tax, environment and data

Netflix has been examined by tax authorities in several jurisdictions over arrangements routing international revenue through the Netherlands, and has restructured to book revenue locally in a number of markets following legislative and public pressure. France, Italy and the United Kingdom have each imposed levies or investment quotas on streaming revenue.

The energy consumption of video streaming has been the subject of contested estimates, some of which were later revised down substantially. Netflix has committed to net zero emissions and reports on production and delivery footprints, with the delivery portion largely determined by the efficiency of networks and devices rather than by the company.

Privacy questions have accompanied the recommendation system, most sharply in the de-anonymisation of the Netflix Prize dataset and a subsequent settlement, and periodically in the company's use of viewing data in marketing, including a 2017 tweet mocking subscribers who had repeatedly watched a particular film, which prompted a discussion about what the company knows.

The streaming competition

The period from 2019 to 2024 is generally described as the streaming wars, and it was a broad industry reorganisation in which almost every participant except Netflix lost a great deal of money.

Disney

The Walt Disney Company was Netflix's most serious challenger and the one with the strongest hand. It owned Marvel, Lucasfilm, Pixar, the Disney animated library, Fox after 2019 and, through its majority in Hulu, an established American streaming business. Disney+ launched in November 2019 at a deliberately low price and reached 100 million subscribers in sixteen months, faster than Netflix had managed in a decade.

It then discovered what Netflix had discovered earlier: that subscriber growth purchased with underpricing does not by itself produce profit. Disney's direct-to-consumer division lost well over US$10 billion cumulatively, its share price fell substantially, Bob Chapek was removed and Bob Iger returned in November 2022 to cut content spending, raise prices, introduce advertising and pursue password sharing, which is to say to do what Netflix had done. Disney's streaming business reached profitability in 2024.

The strategic difference is instructive. Disney entered streaming to protect assets whose existing distribution was declining, which meant its streaming success cannibalised more profitable revenue. Netflix had nothing to protect.

Amazon, Apple and the rest

Amazon operates Prime Video as a component of Prime rather than as a business in itself, which means it does not need to be profitable and cannot easily be measured. It has spent very heavily, most conspicuously around US$1 billion on the first season of The Rings of Power, acquired MGM in 2022, holds NFL Thursday night rights, and introduced advertising by default in 2024, adding an enormous advertising audience overnight.

Apple launched Apple TV+ in November 2019 with no library at all, a very small catalogue and a strategy of prestige rather than volume. It won the Best Picture Academy Award for CODA in 2022, the first streamer to do so, and has produced Ted Lasso, Severance and Slow Horses while remaining, by most estimates, a small and heavily loss-making service that Apple treats as a feature of its device ecosystem.

Warner Bros. Discovery combined HBO Max and Discovery+ under a management that reduced content spending sharply, removed completed programmes from the service for tax write-downs, renamed the service Max and then renamed it HBO Max again, and ultimately agreed in February 2026 to sell itself to Paramount Skydance for US$111 billion after Netflix had initially won the auction for its studios and streaming assets.

Comcast's Peacock and Paramount's Paramount+ have each struggled to reach a scale at which their content spending is defensible, and both have pursued bundling and sport, Peacock with the Olympics and the NBA and Paramount+ with the NFL and Champions League.

The consolidation that followed was predictable from the economics: a market in which content is a fixed cost rewards the largest subscriber base, and there is not room for eight of them.

YouTube

The competitor Netflix has been most careful about in recent statements is not another subscription service but YouTube, which by several measures accounts for a larger share of television viewing time in the United States than Netflix does. Its content costs it nothing to commission, its creators are paid from advertising revenue it shares, and its catalogue is effectively unlimited.

Netflix's response has been to argue that it competes for a different kind of attention, while simultaneously moving into short vertical video with Clips in 2026, into video podcasts from late 2025, and into creator-adjacent programming. The tension between those two positions has not been resolved.

Regional operations

Korea

Netflix launched in South Korea in January 2016 and the market has returned more than any other relative to investment. Kingdom (2019) was the first Korean original, and Squid Game (2021) became the most-watched title in the company's history. Netflix committed US$2.5 billion to Korean content over four years in 2023.

The relationship has not been uncomplicated. Korean broadcasters and producers have argued that Netflix's flat-fee commissioning model captures the upside of global success, as Squid Game demonstrated when its creator received no share of a return estimated at more than forty times the production cost. Netflix has since introduced success-based payments in some Korean agreements. A long dispute with the internet provider SK Broadband over network usage fees, decided partly against Netflix, also fed a wider Korean legislative debate about whether large content providers should pay carriers.

India

India was expected to be Netflix's largest growth market and has been its most difficult. The company launched in 2016 at pricing appropriate to the United States in a market where Disney's Hotstar bundled cricket at a fraction of the cost and Amazon Prime Video was effectively free with delivery.

Netflix cut prices by up to sixty per cent in 2021 and introduced a mobile-only plan, after which subscriptions grew substantially. Its Indian originals, including Sacred Games, Delhi Crime, which won an International Emmy, and Heeramandi, have been well received, and the company committed to substantial local investment, including between ₹60 billion and ₹65 billion announced in February 2024 for the period to 2025.

Content regulation has been a recurring difficulty. India has no pre-certification regime for streaming, but political complaints about A Suitable Boy and Tandav led to police cases and to an industry self-regulation code, and Netflix has made edits in response to objections.

Japan and the rest of Asia

Japan, launched in September 2015, has been strategically important because of anime. Netflix has invested heavily in original anime and in agreements with studios including Production I.G, Bones and Studio Colorido, and has treated anime as a global rather than a Japanese category, which has been vindicated by its performance in Latin America and the Middle East. Live-action Japanese originals including Alice in Borderland and First Love have also travelled. The 2026 World Baseball Classic agreement is the company's first live sport in the country.

Netflix does not operate in mainland China. After several years of attempting entry it licensed individual titles to iQiyi, an arrangement that produced little and has effectively lapsed. Southeast Asia has grown substantially on mobile pricing, and Netflix has commissioned in Thailand, Indonesia and the Philippines.

Europe

Europe is Netflix's largest region outside North America by revenue. The Audiovisual Media Services Directive, amended in 2018, requires that at least thirty per cent of a streaming catalogue in the European Union consist of European works and be given prominence, and several member states impose investment obligations: France requires streamers to spend a proportion of their local revenue on French production and applied a theatrical window that Netflix negotiated down from 36 months to 15. Italy, Spain, Germany and Denmark have comparable schemes.

The effect has been to make Netflix one of the largest commissioners of European television, which producers generally welcome and which some ministries regard with unease on the grounds that a single American buyer now shapes what a national industry makes.

Latin America and Africa

Latin America was Netflix's first large international expansion in 2011 and the region where price sensitivity was learned. Originals including Narcos, Club de Cuervos, La Casa de las Flores and Roma established the company's production presence, and Mexico and Brazil are now substantial production centres.

Africa has been a later and smaller effort, concentrated on Nigeria, where the Nollywood industry produces at very high volume, and on South Africa, with Blood and Water and Queen Sono among the first originals. Connectivity and payment infrastructure have constrained growth more than demand has.

Netflix Games

Netflix announced its entry into video games in July 2021 and launched the service in November of that year, included with every subscription at no additional cost, with no advertising and no in-app purchases.

The reasoning was that games compete for the same attention as television, that the company's existing intellectual property could support them, and that a differentiated offer would reduce cancellations. The commercial reasoning was less obvious, since a feature included free in an existing subscription produces no measurable revenue and can only be justified by retention.

Netflix acquired Night School Studio, the maker of Oxenfree, in September 2021; Next Games, a Finnish mobile developer, in March 2022 for about €65 million; Boss Fight Entertainment in the same month; and Spry Fox in 2022. It opened internal studios in Helsinki and Southern California, the latter under the former Overwatch executive producer Chacko Sonny, and hired Mike Verdu from Facebook Reality Labs to lead the effort.

Results have been mixed. Adoption rates were reported to be very low relative to the subscriber base, with a small fraction of members playing in a given month. The most successful titles have been adaptations of Netflix properties, including Stranger Things games and Squid Game: Unleashed, and licensed mobile ports such as Grand Theft Auto: The Trilogy, which briefly drove very large download numbers. In 2024 the company closed its Southern California studio before it had shipped a title and cancelled a planned triple-A effort, and Verdu moved to a generative artificial intelligence role before leaving.

The strategy has since narrowed to party games playable on a television using a phone as a controller, mobile titles tied to Netflix series, and a small number of licensed games, with the company describing a focus on social play in the living room rather than on competing with dedicated games platforms.

Brand, marketing and consumer products

The Netflix wordmark in its current form dates from 2014 and the stylised N from 2016. The "ta-dum" sound that plays at the start of every title, created in 2015 and refined over months, is among the most recognised audio logos in the world; Netflix has said it was chosen to condition an association between the sound and the beginning of leisure, and an extended orchestral version plays before its theatrical releases.

Marketing has moved substantially away from conventional advertising toward the service itself. The recommendation system is the company's principal marketing channel, since the most effective promotion of a Netflix title is placement on the Netflix home screen, and the personalised artwork system is an advertising apparatus pointed inward.

Tudum, named for the sound, began as an editorial website in 2019, was shut down in the 2022 reductions, and has since returned both as a site and as a live global fan event held in São Paulo in 2020 and 2023 and subsequently elsewhere, at which the company announces programming in the manner of a games publisher's showcase.

Consumer products have grown from a modest licensing operation into a significant one, built principally on Stranger Things, Squid Game, Wednesday and Bridgerton, with collaborations across clothing, toys, food and cosmetics. The Netflix House venues announced in June 2024 extend this into physical experience, combining retail, restaurants and ticketed interactive attractions: Philadelphia with Wednesday and One Piece experiences, Dallas with Stranger Things: Escape the Dark and Squid Game: Survive the Trials, and Las Vegas from 2027. Earlier touring experiences, including the Bridgerton Queen's Ball and Stranger Things: The Experience, tested the format.

Live theatre has followed the same logic, with Stranger Things: The First Shadow in London and on Broadway and Patriots with Peter Morgan in February 2024, both with Sonia Friedman as lead producer.

Music has become an incidental business of some size. Netflix soundtracks have repeatedly reached chart positions, most dramatically the songs from KPop Demon Hunters in 2025, and the company's use of older recordings, notably Kate Bush's "Running Up That Hill" after its appearance in Stranger Things in 2022, has demonstrated an ability to move the music market that no other single outlet now has.

Regulation and public policy

Netflix's scale has drawn regulation of several distinct kinds.

Content quotas and investment obligations in the European Union, France, Italy, Spain, Canada and Australia require streamers to carry or fund local production. Netflix has generally complied and has argued that it would have invested locally regardless, which is largely true and does not dispose of the sovereignty argument underlying the rules.

Network usage fees have been demanded by telecommunications operators in Korea and proposed in the European Union, on the argument that video traffic imposes costs on networks that content providers should share. Netflix's answer is Open Connect, which it offers free and which eliminates most of the cost in question, and its position is that a fee would amount to charging twice for traffic subscribers have already paid to receive. The Korean litigation with SK Broadband went partly against the company before being settled alongside a commercial partnership.

Classification and censorship vary widely. Netflix applies local ratings, and has removed titles in individual markets on legal demand roughly a dozen times, including an episode of Patriot Act with Hasan Minhaj critical of the Saudi government in 2019, material in Singapore, Vietnam and New Zealand, and edits in India. The company publishes these removals, which is more transparency than most of its peers offer and has not prevented criticism that a company advertising creative freedom accepts state instruction where the market is valuable.

Taxation has been addressed by digital services taxes and by streaming levies in France, Italy, Canada and elsewhere, and Netflix has restructured to book revenue in local subsidiaries in a number of countries after criticism of arrangements running through the Netherlands.

Competition scrutiny has been limited because Netflix has made few acquisitions, but the attempted purchase of Warner Bros. Discovery's studios in 2025 would have been the largest test of whether regulators regard streaming as a market in which Netflix holds excessive share, and the argument that the deal carried heavier regulatory risk was a stated reason for WBD's board preferring the Paramount Skydance offer.

Production operations

Netflix's production organisation is now larger than that of most traditional studios and is structured differently.

The company operates production hubs at Albuquerque, New Mexico, where it bought ABQ Studios in 2018 and has expanded repeatedly; at Shepperton in the United Kingdom, under a long lease; in Madrid, its first European production hub, opened in 2019; in Toronto, Seoul, Mumbai, Mexico City, São Paulo, Paris, Berlin, Rome, Stockholm and Tokyo; and at a large new facility at Fort Monmouth, New Jersey, announced in 2022 and under construction.

The organisational distinction is that Netflix largely commissions rather than produces, retaining full rights in exchange for paying the full cost plus a margin upfront. A traditional studio deficit-finances a series, taking a loss on the initial broadcast and recovering it through syndication and international sales. Netflix pays everything at the start, which is attractive to producers who want certainty, and keeps everything afterwards, which is unattractive to producers whose series succeeds. The model has been modified under union pressure and in specific markets, but remains the default.

Post-production, localisation and quality control are handled through a vendor network that Netflix has substantially professionalised, publishing technical specifications, a partner certification programme and style guides for subtitling in dozens of languages. It dubs at a scale no previous distributor attempted, and has been credited with expanding the dubbing industry substantially while also being criticised for rates and for the quality of some early work. Its decision to default new subscribers to dubbed audio in some markets, rather than subtitles, was based on evidence that completion rates were higher, and was resented by parts of the audience.

Netflix's technical specification for original production, which for years required capture at 4K resolution on an approved camera list, had the effect of standardising acquisition formats across the industry and of excluding certain widely used cameras from Netflix work, a commercial consequence for manufacturers that few other single buyers could impose.

Measurement and transparency

For most of its history Netflix published nothing about how much anything was watched, which was a deliberate competitive position: without ratings, rivals could not tell which of its bets had failed, producers could not price their next negotiation, and the press could not run the weekly comparison that structures coverage of broadcast television.

Pressure accumulated from three directions. Producers and unions wanted a basis for compensation. Advertisers on the new tier required audited figures. And competitors began publishing selectively, which made silence look defensive.

The company's definitions changed as the pressure grew. A view was initially counted at two minutes of play, which critics noted counted an abandoned start. From 2021 it reported total hours viewed, then hours divided by runtime, which favours long series. Since December 2023 it has published What We Watched twice yearly, listing every title with more than 50,000 hours of viewing, which is a very large disclosure by the standards of the industry and still does not permit the comparison of a title's audience with its cost.

Nielsen and other third parties now measure streaming in the United States through panels and automatic content recognition, producing figures that broadly correlate with Netflix's own and that occasionally differ enough to be newsworthy. The 2023 union settlements require Netflix to provide viewership data to the guilds under confidentiality, which for the first time gives an outside body a verified view.

The decision announced in 2024 to stop reporting quarterly membership numbers from 2025 runs against this trend and was criticised accordingly, though the company's argument that a membership count means little when memberships cost between three and twenty-five dollars and some carry advertising is not unreasonable.

Timeline

• 1997 Founded on 29 August by Reed Hastings and Marc Randolph in Scotts Valley, California.
• 1998 Netflix.com opens on 14 April with 925 titles and a pay-per-rental model.
• 1999 Introduces the flat monthly subscription with no late fees.
• 2000 Offers itself to Blockbuster for US$50 million and is refused.
• 2002 Lists on Nasdaq on 23 May at US$15 a share.
• 2006 Announces the US$1 million Netflix Prize for recommendation accuracy.
• 2007 Launches streaming as Watch Instantly in January; ships its billionth DVD.
• 2010 Launches in Canada, its first international market; separates streaming and DVD plans.
• 2011 Announces and then cancels Qwikster; loses 800,000 subscribers; the share price falls by three quarters.
• 2013 Releases House of Cards on 1 February, all episodes at once; wins the first streaming Primetime Emmys in September.
• 2015 Releases Beasts of No Nation day-and-date, provoking a cinema boycott; launches in Japan.
• 2016 Launches in 130 further countries on 6 January, reaching more than 190; releases Stranger Things and The Crown.
• 2017 Ted Sarandos-led content spending passes US$6 billion; Okja and The Meyerowitz Stories cause a rule change at Cannes.
• 2019 Roma wins three Academy Awards; Disney+ and Apple TV+ launch.
• 2020 Adds more than 25 million subscribers in the first half of the pandemic; Sarandos becomes co-chief executive.
• 2021 Passes 200 million members; Squid Game becomes the most-watched title in company history; Netflix Games launches.
• 2022 Reports its first subscriber loss in a decade; announces an advertising tier and a crackdown on password sharing; free cash flow turns positive.
• 2023 Hastings becomes executive chairman and Greg Peters co-chief executive; the DVD service ends on 29 September after five billion shipments; the WGA and SAG-AFTRA strikes reshape streaming compensation.
• 2024 Agrees the WWE Raw deal; streams the Paul-Tyson boxing event and its first NFL games; announces Netflix House.
• 2025 Passes 300 million members; KPop Demon Hunters becomes its most popular film; wins and then loses the auction for Warner Bros. Discovery's studios.
• 2026 Receives a US$2.8 billion break fee; introduces Clips; Hastings leaves the board in June.

Financial history

Netflix's financial record divides into four distinct phases, and the transitions between them explain most of the volatility in its share price.

The DVD phase, 1998 to 2010. Revenue grew from nothing to approximately US$2.2 billion, and the business was genuinely profitable on a cash basis because a disc, once bought, could be rented many times and the marginal cost of a rental was postage. Subscribers grew from 239,000 in 2000 to about 20 million in 2010. The company was cash-generative and modestly valued.

The transition, 2011 to 2015. Streaming licences were expensive, international launches consumed capital, and the Qwikster episode destroyed three-quarters of the share price in months. Revenue passed US$3 billion in 2011 and US$6.8 billion in 2015, while net income fell close to zero in 2012 as the company absorbed content costs. Membership passed 40 million in 2013 and 70 million in 2015.

The content build, 2016 to 2021. Revenue grew from US$8.8 billion to US$29.7 billion and membership from 89 million to 222 million. Content spending rose from about US$6 billion to nearly US$17 billion, and free cash flow was persistently and deeply negative, reaching approximately minus US$3.3 billion in 2019. Netflix financed this with high-yield debt totalling more than US$15 billion. The bear case throughout was that the company had never funded its own content and might never be able to. The bull case was that the negative cash flow was an investment in a permanent library and would reverse when growth slowed, which is what happened.

The harvest, 2022 onward. The subscriber loss of early 2022 forced a change of posture. Content spending was held roughly flat rather than growing, advertising was introduced, password sharing was monetised, and prices were raised. Free cash flow turned positive and has remained so. Revenue reached approximately US$31.6 billion in 2022, US$33.7 billion in 2023 and about US$39 billion in 2024, with operating margin rising from roughly 18 per cent to the high twenties. Membership passed 300 million in January 2025. The company began buying back shares substantially and reduced net debt.

The share price history is correspondingly dramatic. Adjusted for two splits, a seven-for-one in 2015 and a ten-for-one in 2025, Netflix has been among the best-performing large American stocks since its 2002 listing, with intervening declines of roughly 80 per cent in 2011 and roughly 75 per cent in 2022, each of which was read at the time as the end of the story.

The DVD operation in detail

The mail-order business that Netflix ran for 25 years is now largely forgotten and was an achievement in its own right.

At its peak around 2010 the service held approximately 100,000 titles and shipped around two million discs a day from more than fifty distribution centres. The centres were deliberately low-technology: discs were sorted by hand, because return flows are unpredictable and because the sleeves are easier for a person to handle than a machine. Staff worked early morning shifts timed to the postal collection, and the operational objective was that a disc returned on Monday should be back in another subscriber's hands on Wednesday.

Netflix worked with the United States Postal Service on the envelope itself through many revisions, optimising for machine sorting, for protection of the disc, and for the reverse fold that let a subscriber return it without a second envelope. The company became one of the Postal Service's largest first-class customers, and the relationship was close enough that changes to postal automation were tested against Netflix mailers.

The queue was the interface. Subscribers ordered the titles they wanted, Netflix sent the top available ones, and the system smoothed demand so that a new release nobody could get did not cause cancellations, since something else always arrived. The practice of throttling, in which heavy users were reported to receive slower service than light ones, produced a class action settled in 2006 with free upgrades, and the company subsequently disclosed the practice in its terms.

The business was profitable until the end. When Netflix announced the closure in April 2023 it still had approximately one million subscribers, largely people in rural areas with poor broadband and film enthusiasts who wanted the catalogue depth that no streaming service has ever matched, since streaming rights for older and less popular films are frequently unavailable at any price. The final shipments went out on 29 September 2023, with subscribers told to keep the discs they held, and the company sent out a number of random additional titles as a farewell. Over its life the service shipped more than five billion discs.

The loss of the DVD catalogue is the clearest illustration of a general point about streaming: the internet made far more content instantly available and made a large body of older work harder to see than it had been in 2005.

Notable programming

Netflix's catalogue is too large to enumerate, but a selection indicates its range.

Drama series include House of Cards, Orange Is the New Black, Narcos, Stranger Things, The Crown, Ozark, Mindhunter, The Haunting of Hill House, Bridgerton, The Queen's Gambit, Wednesday, The Witcher, Dahmer, Beef, Baby Reindeer, One Piece and 3 Body Problem.

Non-English series include Squid Game, Kingdom and The Glory from Korea; Money Heist, Élite and Berlin from Spain; Dark and Kleo from Germany; Lupin from France; Sacred Games, Delhi Crime and Heeramandi from India; Alice in Borderland from Japan; and Narcos: Mexico and La Casa de las Flores from Mexico.

Comedy includes Arrested Development, BoJack Horseman, Master of None, Russian Doll, The Kominsky Method, Never Have I Ever, Sex Education and Nobody Wants This, alongside an enormous catalogue of stand-up specials from Dave Chappelle, Chris Rock, Ali Wong, Hannah Gadsby, Bo Burnham and many others.

Unscripted includes Making a Murderer, Tiger King, Love Is Blind, The Great British Baking Show in the United States, Drive to Survive, Queer Eye, Nailed It!, Old Enough and Physical: 100.

Documentary includes 13th, Icarus, American Factory, My Octopus Teacher, Our Planet, The Social Dilemma and Don't F**k with Cats.

Animation includes BoJack Horseman, Arcane, Klaus, The Sea Beast, Guillermo del Toro's Pinocchio, which won the Academy Award for Best Animated Feature in 2023, Nimona and KPop Demon Hunters.

Film includes Beasts of No Nation, Roma, The Irishman, Marriage Story, The Trial of the Chicago 7, The Power of the Dog, All Quiet on the Western Front, Glass Onion, Maestro, Emilia Pérez and the Red Notice and Extraction action franchises.

Awards and recognition

Netflix received its first major Primetime Emmy Awards in September 2013 for House of Cards, the first for a programme distributed online, and has since been among the leading nominees each year, overtaking HBO in total nominations in 2018 and winning the most Emmys outright in several subsequent years.

Its films have won Academy Awards including Best Documentary Feature for Icarus (2018), Free Solo having gone elsewhere, American Factory (2020) and My Octopus Teacher (2021); Best Supporting Actress for Laura Dern in Marriage Story (2020); Best Director and Best Cinematography for Alfonso Cuarón for Roma (2019), which also won Best International Feature; Best Director for Jane Campion for The Power of the Dog (2022); and four awards for All Quiet on the Western Front (2023). The Best Picture award has not yet gone to a Netflix film, a gap the company has pursued with expensive campaigns and which is generally attributed to residual resistance within the Academy to a distributor that does not respect the theatrical window.

Squid Game won six Emmys in 2022, including Outstanding Lead Actor for Lee Jung-jae, the first for a non-English-language performance, and Outstanding Directing for Hwang Dong-hyuk.

Leadership

Reed Hastings was chief executive from 1999 to 2020, co-chief executive to January 2023, executive chairman thereafter and left the board in June 2026. He has since concentrated on philanthropy in education, where he has been a long-standing and sometimes controversial funder of charter schools, and on his ownership of the Powder Mountain ski resort in Utah.

Marc Randolph was the first chief executive and left in 2003. He has written about the founding in That Will Never Work (2019) and has been a persistent corrective to tidied-up accounts of the company's origins.

Ted Sarandos joined in 2000 from Video City and ETD and built the content organisation from licensing to global commissioning. He became chief content officer, then co-chief executive in July 2020, and is the executive most associated with the company's creative strategy.

Greg Peters joined in 2008, led the international expansion and device partnerships, served as chief operating and product officer, and became co-chief executive in January 2023. He is generally identified with the advertising tier, paid sharing and the product side of the business.

Other senior figures have included Spencer Neumann as chief financial officer, Elizabeth Stone as chief technology officer, Eunice Kim as chief product officer, Bela Bajaria as chief content officer, and Patty McCord, whose work on the culture document in the early years shaped the company's reputation as an employer.

The board has included Jay Hoag, Rodolphe Belmer, Mathias Döpfner, Timothy Haley, Leslie Kilgore, Strive Masiyiwa, Ann Mather, Susan Rice, Brad Smith and Anne Sweeney, among others.

Effect on the form of television

Netflix changed not only how television is distributed but how it is written, and the changes are visible enough that critics have given them names.

The season released at once removed the week between episodes, and with it the recap, the cliffhanger designed to sustain seven days of speculation, and the communal experience of an audience watching simultaneously. Writers adapted by treating a season as a single long film, which suits some stories and starves others of the accumulation that weekly release produces. Several of the most discussed series of the period, including Succession and The White Lotus on HBO and Severance on Apple, released weekly and generated a conversational presence that Netflix titles, watched and finished within a weekend, often did not. Netflix has responded by splitting seasons into two parts for its largest titles, including Stranger Things, Bridgerton and Wednesday, which is a partial return to scheduling.

The shorter season followed from the absence of an advertising schedule to fill. Eight or ten episodes replaced twenty-two, which raised the average quality of an episode and reduced the total employment the industry provided, and which removed the filler episode that in practice was often where a long-running series developed its minor characters.

The second-screen problem is the most discussed recent criticism. Producers have reported instructions from streaming buyers to have characters state what they are doing so that a viewer looking at a phone can follow, and screenwriters have described notes asking for dialogue to be more explicit for the same reason. Netflix has denied issuing such guidance as policy. Whether or not it is formalised, the pattern is widely enough reported by working writers to be taken seriously, and it follows logically from optimising for completion rate.

The algorithm's shaping of commissions has been argued in both directions. One view holds that data-driven buying produces derivative work, since the evidence available always concerns what already exists. The opposing view, which the record supports at least partly, notes that Squid Game, Stranger Things, Making a Murderer and KPop Demon Hunters were not obvious commissions by any prior standard, and that a company with global distribution can afford idiosyncratic bets precisely because a small share of 300 million subscribers is a large audience.

The theatrical question

Netflix's relationship with cinema has been adversarial since 2015 and the dispute has never been settled, only negotiated around.

The exhibition industry's position is that the theatrical window, in which a film plays in cinemas exclusively for a period before any home release, is what makes cinema-going worth the effort, and that a film available at home on the day of release will not be seen in a cinema. Netflix's position is that its subscribers have paid for access, that withholding a film from them for three months in order to benefit a separate industry is indefensible, and that the window is a convention rather than a law.

The consequences have been concrete. AMC, Regal and Cineworld have generally refused Netflix films, so the company releases through independent cinemas in limited runs, principally to qualify for awards. Roma had a three-week window; The Irishman a limited run; Glass Onion an unusual one-week wide release in 2022 that grossed a reported US$15 million and which Netflix declined to extend despite exhibitor requests, a decision widely read as demonstrating that it could succeed theatrically and chose not to.

The Cannes Film Festival changed its competition rules in 2018 to require French theatrical release, which under French law imposes a long window, and Netflix withdrew from competition. Okja and The Meyerowitz Stories had competed in 2017 and been booed at their screenings by some French exhibitors.

The Academy debated a rule change after Roma. Steven Spielberg argued that a film released principally on a streaming service was a television film and should compete at the Emmys; Alfonso Cuarón and others replied that a rule requiring theatrical release would exclude films that cinemas refused to show. The Academy did not change the rule materially, and in 2022 Apple became the first streamer to win Best Picture, with CODA.

Under Dan Lin, appointed film chief in 2024, Netflix has moved somewhat, committing to longer and wider theatrical runs for selected titles while maintaining that the window remains a matter for it to decide.

Workforce and offices

Netflix employs roughly 14,000 people, a strikingly small number for a company of its revenue and one that reflects both the culture's preference for a small number of highly paid people and the fact that production crews are employed by productions rather than by the company.

Compensation is paid in cash at the top of the market, with employees able to elect what proportion to take as stock options, an unusual arrangement that removes the retention effect of vesting and is consistent with the culture's insistence that people should stay because they want to.

The headquarters is at Los Gatos, California, in a campus of buildings named after films. A second major American site at Hollywood, in the Epic and Icon towers on Sunset Boulevard, houses the content organisation, and the company has substantial offices in Amsterdam, which serves as the European headquarters, in London, Singapore, Tokyo, Seoul, Mumbai, Mexico City, São Paulo, Madrid, Paris, Berlin, Rome, Sydney and Toronto.

The parental leave policy announced in 2015, offering salaried employees unlimited leave in the first year after a birth or adoption, received considerable attention and was later modified; it initially did not extend to hourly workers in the DVD operation, which drew criticism.

Philanthropy and social commitments

Netflix announced in 2020 that it would move two per cent of its cash holdings, initially around US$100 million, into financial institutions serving Black communities in the United States, and later extended the commitment. It established the Netflix Fund for Creative Equity in 2021, committing US$100 million over five years to training and placing under-represented people in the screen industries, and commissioned research from the University of Southern California's Annenberg Inclusion Initiative into the diversity of its own output, publishing the results including where they were unflattering. The 2021 and subsequent reports found Netflix ahead of the industry on gender and on some measures of racial representation and behind on disability.

The company operates the Netflix Hardship Fund for crews affected by production shutdowns, established with US$150 million during the pandemic, and has funded creative equity programmes in several countries, including a substantial commitment in Nigeria and a fund for European production talent.

Hastings's own philanthropy, conducted separately, has concentrated on education, particularly charter schools, and he and his wife Patty Quillin gave US$120 million to historically Black colleges and universities in 2020. He has also given large sums to Democratic candidates and causes, and his educational giving has been criticised by opponents of charter expansion.

Assessment

Netflix's significance rests on two decisions taken roughly a decade apart, each of which involved damaging a working business in order to build the one that would replace it.

The first was to invest in streaming from 2007 while the DVD business was profitable and growing, and to keep investing through the Qwikster disaster, which was a botched execution of a correct strategy. The second was to commission original programming from 2011, which converted the company from a distributor dependent on studios into a competitor to them, and which the studios could have prevented by refusing to license anything at all in the years when Netflix still needed them.

The outcome reshaped an industry. Every large media company now operates a streaming service; the scheduled television evening has largely ended in the markets where Netflix operates; subtitled drama from Korea, Spain and Germany reaches global audiences; and the amount of scripted television produced annually roughly doubled during the period of maximum competition before contracting again.

The costs are also clear, and they fall unevenly. The people who make television have less durable income than they did, cinema has lost a category of film, and the audience has gained abundance at the price of a catalogue that changes without notice and series that end without conclusion.

The open question is what Netflix becomes now that the growth phase has ended. Its recent moves, into advertising, live sport, wrestling, podcasts, physical venues, games and short-form video, describe a company assembling the attributes of the broadcasters it displaced. The reported exploration of linear channels and third-party bundles in 2026 makes the resemblance explicit. Whether that represents a mature business extending its franchise sensibly, or the same convergence that overtook every previous dominant distributor, is not yet possible to say, and the failed pursuit of Warner Bros. Discovery suggests the company itself is uncertain which it is.

See also

• Reed Hastings
• Ted Sarandos
• Marc Randolph
• Streaming media
• The Walt Disney Company
• Warner Bros. Discovery
• Amazon
• Blockbuster LLC