The Living Room's Landlord
In the first quarter of 2024, Roku reported that its platform streamed 30.8 billion hours of content — more than Netflix, more than YouTube on connected TVs, more than any single programmer could claim on American television screens. The company that makes this possible earned $881.6 million in revenue that quarter, yet the unit that actually manufactured and sold the hardware — the little boxes and TV sticks that consumers associate with the Roku brand — generated just $151.8 million of it, at a gross loss. The other $729.8 million came from something else entirely: the right to control what 81.6 million active accounts see when they turn on their television. Roku loses money putting screens in living rooms so it can tax the attention that flows through them. This is the defining arithmetic of one of the strangest and most consequential businesses in consumer technology — a company that gives away the razor, the handle, and the shaving cream, then collects rent on the bathroom mirror.
The strangeness runs deeper than a pricing model. Roku is not a streaming service. It is not, in any meaningful sense, a hardware company. It is the operating system of the American living room — the default interface between 120 million people and the $100-billion-plus streaming economy — and it achieved this position not through the overwhelming capital advantages that typically define platform wars, but through a sixteen-year campaign of disciplined self-denial. Every major competitor in the connected TV space — Amazon, Apple, Google, Samsung — operates the television as a loss leader for a larger commercial empire. Roku operates the television as the thing itself, which is simultaneously its greatest vulnerability and the source of an extraordinary structural advantage: content companies will distribute through Roku precisely because Roku doesn't compete with them for subscribers.
The company trades at a market capitalization that fluctuates between $8 billion and $13 billion, depending on the market's mood about streaming economics. This makes it roughly 1/200th the size of Apple and 1/150th the size of Alphabet. And yet when Americans turn on a smart TV in 2024, there is a roughly one-in-three chance the screen that greets them is Roku's. The gap between the company's scale in market capitalization and its scale in consumer attention is the gap this profile exists to explain.
By the Numbers
Roku at a Glance (FY2024)
$4.1BTotal net revenue (FY2024)
$3.47BPlatform revenue (FY2024)
81.6MActive accounts (Q4 2024)
30.8BStreaming hours (Q1 2024)
#1Streaming OS market share in the U.S.
~$44Average revenue per user (trailing 12 months)
~3,500Employees (approx.)
The Reluctant Founder's First Machine
Anthony Wood is the kind of serial entrepreneur Silicon Valley used to produce more reliably — an engineer who kept building things because the things he'd already built made the next thing obvious. Born in England, raised in Texas, he studied electrical engineering at Texas A&M and then did the thing hardware-inclined founders do in the late 1990s: he started a company to make a consumer device that the market hadn't quite decided it wanted yet. That company was ReplayTV, one of the two original digital video recorders alongside TiVo, and its trajectory would shape everything Wood built afterward.
ReplayTV was arguably the superior product — it allowed users to skip commercials automatically and share recordings over the internet, features so threatening to the broadcast establishment that a consortium of entertainment companies sued ReplayTV's parent company into bankruptcy. TiVo survived with a more conciliatory approach to the industry. Wood left before the implosion, but the lesson calcified into something like a founding theology: do not antagonize the content owners. If the DVR wars taught him anything, it was that the companies that create programming possess a veto power over distribution technologies that no amount of engineering elegance can override. Build with them, not against them.
Wood moved on, building and selling another company — SonicBlu's subsidiary — and in 2002 founded Roku. The name means "six" in Japanese, a reference to it being his sixth company. For its first several years, Roku operated in relative obscurity, developing streaming hardware with no particular breakthrough in sight. The inflection came from an unexpected direction.
In 2007, Netflix was preparing to launch its streaming service.
Reed Hastings's team had spent two years building a hardware device — a small set-top box codenamed "Project Griffin" — that would plug into televisions and deliver Netflix content directly. The project was nearly complete when Hastings made a characteristically counterintuitive decision: he killed it. The reasoning was elegant and, in retrospect, obvious. If Netflix manufactured its own hardware, every other device maker would view Netflix as a competitor and deprioritize its app. Hastings wanted Netflix on every screen; owning a screen would make that harder. He spun the hardware team and the nearly finished device out to a small company that had been doing contract work on the project. That company was Roku.
We realized that nobody was trying to be the platform for streaming. Everyone wanted to be a streaming service or sell you hardware that locked you into their ecosystem. We just wanted to be the platform.
The Netflix Player by Roku launched in May 2008 at $99. It was a simple purple box that did one thing well: it connected your television to Netflix's streaming library. Within a year, Roku had expanded beyond Netflix exclusivity — a critical early decision — adding channels for Amazon Video, MLB.tv, and a growing roster of content partners. The Netflix exclusive relationship lasted only months, but the distribution deal gave Roku something more durable: legitimacy, initial scale, and a mental model for what it wanted to become. Not a Netflix accessory. Not a hardware vendor. A neutral platform.
The Switzerland Strategy
Platform neutrality is one of those strategic concepts that sounds obvious in retrospect and nearly suicidal in real time. Through the early 2010s, the connected TV market was being reshaped by companies with essentially unlimited resources. Amazon launched Fire TV in April 2014, backed by the full weight of Prime's ecosystem and Amazon's willingness to sell hardware at or below cost. Apple had the Apple TV, priced at a premium and integrated into the iPhone-iPad-Mac constellation. Google had Chromecast at $35, later Android TV, and the gravitational pull of YouTube. Samsung and LG were building proprietary smart TV platforms — Tizen and webOS — into their sets.
Against this lineup, Roku had no content library, no smartphone ecosystem, no e-commerce flywheel, no semiconductor supply chain, no global manufacturing footprint. What it had was something none of these competitors could credibly offer: the absence of a conflict of interest. When Disney was deciding where to launch Disney+ in November 2019, it needed every major streaming platform. But distributing through Amazon Fire TV meant enriching a company that competed with Disney in content production through Amazon Studios. Distributing through Apple TV meant strengthening a company that sold its own $4.99/month Apple TV+ subscription. Roku wanted none of these things. Roku wanted a share of the ad inventory and a percentage of subscription revenue driven through its platform — a tax, not a competitive threat.
This positioning required ferocious discipline. Wood and his team had to resist the temptation to launch a Roku-branded subscription service, to acquire content libraries, to build the kind of integrated ecosystem that Wall Street rewards with premium multiples. Every strategic decision was filtered through a single question: does this make content partners more or less likely to distribute through us? When Roku eventually launched The Roku Channel in 2017 — a free, ad-supported streaming service — it was designed with deliberate limitations. It aggregated licensed and free content rather than producing expensive originals, and it was made available on competing platforms (Samsung TVs, Amazon Fire TV) precisely to signal that Roku was not trying to build a walled garden.
The neutrality strategy had a second, less visible dimension. Because Roku had no adjacent business to subsidize, it had to build an operating system that worked — that was fast, intuitive, and reliable on extremely cheap hardware. The constraints of survival bred a particular engineering culture obsessed with stripping out complexity. The Roku interface remained stubbornly simple: a grid of apps, a search bar, and not much else. This simplicity became its own advantage. The median American consumer buying a television is not a technologist. They want to press a button and find their show. Roku's interface, designed for $29 streaming sticks and $149 smart TVs, optimized relentlessly for that use case.
The TV [Trojan Horse](/mental-models/trojan-horse)
The most important strategic decision in Roku's history was not to build a better streaming stick. It was to stop caring about the stick altogether.
In 2014, Roku began licensing its operating system to television manufacturers. The logic was disarmingly simple: if the streaming OS was embedded in the television itself, consumers would never need to buy a separate device. They would turn on the TV and land directly in Roku's platform. The first Roku TV, manufactured by TCL and Hisense under license, appeared that year. It was cheap — sub-$200 for a 32-inch set — and it was good enough. Not good enough to threaten Samsung or LG at the high end, but good enough for the bedroom, the dorm room, the first apartment. Good enough for the price-sensitive majority of the American television market.
This was the move that transformed Roku from a device maker into a platform company. By 2024, Roku had become the number-one smart TV operating system in the United States, Canada, and Mexico, powering TVs from TCL, Hisense, Philips, and a growing list of manufacturers. One in three smart TVs sold in the U.S. ran Roku's OS. The company's installed base — those 81.6 million active accounts — was not primarily composed of people who had bought a Roku stick. It was composed of people who had bought a television and happened to get Roku with it.
The genius of the TV licensing model was that it outsourced the capital-intensive, margin-destroying business of consumer electronics manufacturing to companies that already knew how to do it. TCL and Hisense competed on panel price, supply chain efficiency, and retail shelf space. Roku competed on software and advertising monetization. The TV manufacturers got a polished, continuously updated operating system they didn't have to build. Roku got distribution at a scale no direct-to-consumer hardware company could match — every Walmart television aisle in America became a Roku customer acquisition channel, with the TV manufacturer bearing the inventory risk.
We don't make money selling TVs. We don't make money selling players. We make money when people use their TVs, and the more people who use Roku TVs, the more we can make.
In March 2023, Roku went further: it announced its own Roku-branded television line, manufactured in partnership with contract manufacturers. The move seemed to contradict the Switzerland strategy — wouldn't this antagonize TCL and Hisense? But Roku framed it carefully. The Roku-branded TVs were positioned in the ultra-budget segment, below the price points where its licensed partners competed most aggressively. And the rationale was not hardware margin. It was reference design control. By manufacturing its own sets, Roku could dictate every aspect of the user experience — the remote control, the boot-up sequence, the home screen layout — and demonstrate best practices that its licensed partners could then adopt. The Roku TV became the concept car for the platform.
The Economics of [Attention](/mental-models/attention) [Arbitrage](/mental-models/arbitrage)
To understand Roku's business, you have to understand a single number: average revenue per user, or ARPU. In Q4 2024, Roku's trailing twelve-month ARPU stood at approximately $43.63. This means that each of Roku's 81.6 million active accounts generated roughly $44 in annual revenue — revenue derived not from selling those users a device, but from monetizing their presence on the platform through advertising, content distribution agreements, and transaction fees.
The mechanics of platform monetization at Roku operate across three interconnected revenue streams, all classified under the "Platform" segment.
First, advertising. When a Roku user turns on their television, they see Roku's home screen. That home screen has banner ads, sponsored content rows, and screensaver advertisements — all controlled by Roku and sold to advertisers. When users watch ad-supported content through The Roku Channel or through third-party apps that have agreed to share inventory, Roku earns revenue on those impressions. Roku's demand-side platform, built through acquisitions including its 2021 purchase of Nielsen's Advanced Video Advertising business (and later augmented by its own technology), allows advertisers to buy connected TV inventory programmatically.
Second, content distribution fees. When a user subscribes to a streaming service through the Roku platform — signs up for Paramount+, Peacock, or Starz via the Roku interface — Roku takes a share of the subscription revenue, typically reported in the range of 20% to 30% in the first year or two, declining thereafter. This is the app store model applied to television. Roku also negotiates for a share of ad inventory within third-party apps. The standard arrangement historically required content partners to surrender roughly 30% of their ad impressions to Roku's sales team.
Third, transactional revenue. Users who rent or purchase movies and shows through the Roku platform generate take-rate revenue for the company.
The platform segment generated $3.47 billion in revenue in FY2024 at a gross margin around 53%. The device segment — the sticks, the soundbars, the Roku-branded TVs — generated approximately $603 million at a gross margin that fluctuated near or below zero. Roku's device business is, by design, a cost center. Every dollar of margin sacrificed on hardware is an investment in acquiring an active account that will generate $44 per year in platform revenue — recurring, high-margin, and growing.
Roku's revenue model in FY2024
| Segment | Revenue | % of Total | Gross Margin |
|---|
| Platform | $3.47B | ~85% | ~53% |
| Devices | $603M | ~15% | ~0% (breakeven to slight loss) |
| Total | $4.07B | 100% | ~45% |
The ARPU number is the heartbeat metric. It tells you how effectively Roku is monetizing its installed base. And here the trend was encouraging but complicated. ARPU grew from $27.95 in 2020 to $41.03 in 2022 — a surge driven by the pandemic's acceleration of streaming adoption and the connected TV advertising boom. Then it plateaued, dipping slightly in 2023 before recovering in 2024. The plateau reflected two things: a macro slowdown in digital advertising, and the mathematical dilution of adding international accounts (which monetize at a fraction of U.S. rates) to the denominator. The U.S. ARPU, which Roku does not break out separately, is estimated to be substantially higher — perhaps $55 to $65 — obscured by the growing share of lower-monetization international users.
The War for the Home Screen
The battle for the connected TV operating system is one of the most consequential and least understood competitive wars in technology. It is, in essence, a replay of the smartphone platform war — Android versus iOS, with the television as the contested territory — except the economic stakes are different and the number of combatants is larger.
In the United States, four operating systems control the vast majority of smart TVs and streaming devices: Roku OS, Amazon Fire TV, Google's Android TV / Google TV, and Samsung's Tizen. Apple's tvOS commands a smaller but high-value segment. LG's webOS has meaningful share in premium TVs. Vizio's SmartCast, before Walmart's acquisition of Vizio in 2024, occupied a niche similar to Roku's strategy in certain demographics.
Roku's claim to the number-one position in U.S. streaming is based on reach: by the company's own reporting, Roku devices accounted for the largest share of streaming hours and the largest installed base of active accounts in the United States. Third-party data from Parks Associates, Conviva, and others broadly confirmed this, though market share estimates varied depending on whether you measured by device shipments, installed base, or streaming hours. By most measures, Roku held between 30% and 38% of the U.S. connected TV market as of 2024, with Amazon Fire TV as the closest competitor at roughly 28% to 32%.
The competitive dynamics are brutal. Amazon can sell Fire TV sticks at $24.99 because it profits when users buy things on Amazon, subscribe to Prime Video, and stay inside the Amazon ecosystem. Google can subsidize Android TV and Chromecast because connected TVs feed Google's advertising machine and YouTube's dominance. Samsung bundles Tizen into every Samsung TV — the world's largest TV brand by unit volume — meaning its share grows automatically with TV sales. Apple charges premium prices and earns subscription revenue from Apple TV+, Apple Music, and its services ecosystem.
Roku's competitive response has been, characteristically, to go lower. In late 2023, Roku launched a line of smart TVs at Walmart starting at $148 for a 55-inch 4K set — a price point that seemed to defy the economics of television manufacturing. The company was willing to lose money per unit sold, banking on the lifetime platform revenue each TV would generate. This is the connected TV equivalent of Amazon's Kindle strategy: the device is a vessel for monetizable engagement.
The home screen is the most valuable real estate in media. Whoever controls the home screen controls what gets watched.
But the competitive landscape shifted ominously in 2024. Walmart's $2.3 billion acquisition of Vizio, completed in December 2024, was explicitly motivated by access to Vizio's SmartCast advertising platform and its 18 million active accounts. Walmart now owned a connected TV OS and could integrate it with the largest physical retailer's advertising and data capabilities. The deal signaled that the connected TV platform was becoming valuable enough to attract retail conglomerates — and that Roku's competitive set was expanding beyond technology companies into the broader advertising-industrial complex.
Amazon, meanwhile, was deepening Fire TV's integration with its advertising business. In 2024, Amazon began inserting ads into Prime Video by default — a move that expanded its CTV advertising inventory dramatically and put direct pressure on Roku's ad sales. Google continued to push Google TV into more OEM partnerships globally. The encirclement was real.
The Roku Channel and the Content Paradox
The Roku Channel, launched in September 2017, began as a modest aggregation play: a free, ad-supported destination offering licensed movies and TV shows. It was, in effect, Roku's answer to a strategic problem. As the platform grew, Roku needed to demonstrate to advertisers that it could control and sell premium ad inventory at scale. Third-party apps like Netflix and Disney+ either had no ads or sold their own inventory. The Roku Channel gave Roku a first-party content surface where it controlled 100% of the ad load.
By 2024, The Roku Channel had grown into one of the largest free ad-supported streaming television (FAST) services in the United States. It offered over 350 linear channels, a library of on-demand movies and TV shows, and, increasingly, original programming. Roku City — the animated screensaver that appears when a Roku device is idle — had become a cultural artifact, a piece of ambient television that generated its own ad impressions. The Roku Channel reached U.S. households in the tens of millions, making it a top-five FAST service alongside Tubi (owned by Fox), Pluto TV (owned by Paramount Global), and Amazon's Freevee (which Amazon folded into Prime Video in 2024).
But The Roku Channel created a paradox that Wood and his team had to navigate carefully. Every dollar invested in original content for The Roku Channel risked the neutrality positioning that had been Roku's strategic foundation for sixteen years. If Roku spent $500 million a year on original programming — as some investors urged during the streaming content arms race of 2021–2022 — it would become a competitor to the very services it hosted. Content partners would have reason to deprioritize Roku's platform or negotiate harder on revenue-sharing terms.
Roku's resolution was characteristically restrained. It invested in originals, but modestly — acquiring the library of the defunct Quibi platform for a reported pittance in January 2021, licensing rather than producing most content, and keeping original content budgets well below those of any major streamer. The Roku Channel's content strategy was FAST-first: aggregating free linear channels, licensing second-run and library content, and using original programming as accent pieces rather than centerpieces. The discipline held, though whether it could hold indefinitely — as The Roku Channel grew larger and the temptation to invest more in content intensified — remained an open question.
The Ad Tech Transformation
The real Roku story — the one that determines whether the company is worth $10 billion or $50 billion in a decade — is the advertising story. And the advertising story is a story about the migration of television ad dollars from linear to streaming, and Roku's bet that it can capture a disproportionate share of that migration.
The U.S. television advertising market has historically been a $70 billion annual category. Linear TV — broadcast and cable — still commanded the majority of that spending as of 2024, but the trajectory was unmistakable. Linear TV viewership was declining 5% to 8% annually. Streaming viewership was growing. And yet advertising budgets, governed by institutional inertia, longstanding agency relationships, and the mechanical complexity of buying CTV ads, had not migrated at anywhere near the rate of eyeballs. This gap — the gap between where attention had moved and where ad dollars still sat — was the core market opportunity for every connected TV platform.
Roku attacked it through a series of acquisitions and product investments. In April 2021, Roku acquired Nielsen's Advanced Video Advertising business for an undisclosed sum, gaining capabilities in automatic content recognition (ACR) technology — the ability to identify what a viewer is watching on their Roku device, whether through a streaming app, a cable box, or an antenna. ACR data is the connective tissue of CTV advertising: it allows Roku to measure cross-platform reach, deduplicate audiences, and prove to advertisers that their spending drove results.
In 2023 and 2024, Roku built out its proprietary demand-side platform and expanded its self-service advertising tools, allowing smaller advertisers — the long tail of local businesses and direct-to-consumer brands — to buy Roku advertising programmatically. The company launched shoppable ads that allowed viewers to purchase products directly from their TV remote, integrating commerce into the advertising experience. Roku also introduced interactive ad formats and began experimenting with AI-driven ad targeting and creative optimization.
The financial thesis was straightforward. If U.S. CTV advertising grew from roughly $25 billion in 2023 to a projected $40 billion to $50 billion by 2027 — as eMarketer and other forecasters estimated — and if Roku maintained or expanded its share of CTV ad spending, platform revenue would continue to compound even without significant growth in active accounts. The shift from hardware company to advertising platform was Roku's metamorphosis, and it was well underway. By 2024, the majority of Roku's platform revenue derived from advertising rather than content distribution fees.
📺
The CTV Advertising Opportunity
U.S. connected TV ad spending trajectory
2019U.S. CTV ad spending: ~$7 billion
2021CTV ad spending surges to ~$14 billion amid pandemic streaming boom
2023CTV ad spending reaches ~$25 billion; Roku platform revenue hits $3.1B
2025EIndustry forecasts project CTV ad spending of $30–35 billion
2027ECTV ad spending projected at $40–50 billion, approaching parity with linear TV
Abby and the Profitability Question
For a company that generated $4.1 billion in revenue in FY2024, Roku's relationship with profitability has been, to use the most generous framing, aspirational. The company reported a net loss of approximately $710 million in FY2023 and was expected to narrow losses significantly in FY2024, with adjusted EBITDA turning positive in certain quarters. But GAAP profitability remained elusive. Stock-based compensation was substantial — in the hundreds of millions annually — and operating expenses, particularly in research and development and sales and marketing, consistently outpaced gross profit growth.
The bear case on Roku has always centered on this gap. The company has been public since September 2017, when it IPO'd at $14 per share and raised $252 million. In the seven years since, it has never posted a full-year GAAP net profit. The stock peaked above $470 in July 2021, during the pandemic streaming euphoria, then collapsed more than 90% to below $40 in late 2022 as interest rates rose and the advertising market contracted. By mid-2024, it traded in the $55 to $70 range — a company that had destroyed enormous shareholder value from peak to trough and still needed to prove it could generate durable free cash flow.
Wood's response to the profitability question was always the same: the company was investing in growth, the CTV market was early, and the time to harvest would come. In 2023, he began a restructuring effort — reducing headcount by approximately 10% in two rounds of layoffs, rationalizing content spending, and tightening operating expense growth. The messaging shifted from "invest at all costs" to "disciplined growth with a path to sustained profitability." By late 2024, the company was guiding toward positive adjusted EBITDA on an annual basis.
The question investors had to answer was whether Roku's cost structure could flex downward faster than its competitors could erode its revenue growth. The advertising business had strong unit economics — selling ads against existing inventory is high-margin work — but the investment required to maintain the platform, negotiate content deals, build ad tech, and subsidize hardware was relentless. Roku's path to profitability required ARPU to keep growing, which required the CTV ad market to keep expanding, which required Roku to maintain its platform share against Amazon, Google, Samsung, and now Walmart-Vizio. Every link in the chain had to hold.
The International Gambit
The United States was Roku's stronghold, but the U.S. smart TV market was approaching saturation — household penetration of connected TVs exceeded 85% by 2024. Growth in active accounts would increasingly need to come from international markets, where Roku's presence was, charitably, nascent.
Roku had expanded into Canada, the UK, Mexico, and several Latin American and European markets through a combination of streaming device sales and OEM partnerships. But international markets presented structural challenges that the U.S. playbook didn't fully address. The TV OEM landscape was different — in Europe, Samsung and LG dominated, and Android TV had deeper penetration through partnerships with brands like Sony, Philips (in some markets), and others. Content licensing was fragmented by geography. Advertising ecosystems were less developed and more heavily regulated. And Roku's ARPU in international markets was a fraction of U.S. levels, meaning each new international account diluted the company-wide ARPU metric that Wall Street watched.
The strategic tension was clear: Roku needed international growth to justify its valuation, but international expansion consumed capital and compressed margins. The company's approach was cautious — expanding primarily through TV OEM partnerships rather than the expensive direct-to-consumer marketing campaigns that had driven U.S. adoption. In 2024, Roku signed deals with additional TV manufacturers for Latin American and European distribution, but international revenue remained a small and low-margin portion of the overall business.
The Culture of the Purple Box
Roku's corporate culture reflected its founder's engineering temperament — frugal, product-obsessed, and allergic to the performative excess of Silicon Valley. The company's headquarters in San Jose, California, was notably unglamorous by tech standards. Wood, who remained CEO and controlled a majority of voting power through a dual-class share structure, ran the company with an engineer's directness. Meetings were short. Strategy memos were expected to be precise. The company's annual all-hands meetings were functional rather than theatrical.
This culture had advantages and costs. On the advantage side, Roku's engineering teams shipped reliably. The OS was updated continuously, with features rolling out across the entire installed base simultaneously — a benefit of controlling the full software stack. Roku's developer relations team maintained relatively frictionless relationships with the thousands of app developers on the platform, owing in part to the simplicity and stability of the Roku developer toolkit.
On the cost side, the frugality sometimes bordered on austerity. Roku's compensation, particularly in base salary and cash bonus, lagged behind the packages offered by Amazon, Google, and Apple for comparable engineering talent. The company relied heavily on stock-based compensation, which worked beautifully when the stock was rising 300% per year in 2020–2021 and became a retention crisis when it fell 90% in 2022. The 2023 layoffs — approximately 600 to 700 employees across two rounds — were partly a response to overhiring during the pandemic boom and partly a necessary recalibration of the company's cost structure.
Wood's dual-class control meant that Roku was, in practice, a founder-controlled company in which short-term shareholder pressure could be — and was — ignored. This gave the company the luxury of long-term investment discipline but also insulated Wood from the kind of accountability that might have accelerated the path to profitability. The tension between founder vision and shareholder value is one of the permanent structural features of Roku's governance.
The Living Room as Marketplace
By late 2024, the contours of Roku's long-term vision had become clearer — and more ambitious than the "neutral streaming platform" framing suggested.
Roku was building the television into a commerce platform. Shoppable ads — where a viewer could press a button on the Roku remote to purchase a product seen in a commercial — were live and growing. Roku Pay, the company's payment infrastructure, processed subscription sign-ups and in-app purchases. The Roku home screen, which greeted users every time they turned on their TV, was being redesigned to surface not just streaming content but interactive experiences, games, and advertising-funded entertainment.
The company was also exploring AI-driven content discovery and personalization, using viewing data from 81.6 million accounts to recommend content across streaming services — a capability that positioned Roku as a super-aggregator, the Google of the television. If a viewer wanted to watch a specific actor's filmography, Roku's universal search could surface results across Netflix, Hulu, Max, Disney+, and The Roku Channel, and direct the viewer to the cheapest or most convenient option. The platform that controlled this search and discovery layer would, over time, exercise enormous influence over which content got watched and which streaming services succeeded.
The TV is becoming the largest screen in the home for everything — not just entertainment, but shopping, fitness, video conferencing. We're building the platform for all of it.
This was the deepest ambition: to make Roku not just the operating system for streaming, but the operating system for the television as a general-purpose device. It was an ambition that placed Roku in competition not just with Amazon and Google but with the entire concept of the television as a passive display. Whether a company of Roku's scale — three to four billion dollars in annual revenue, perpetually unprofitable, competing against trillion-dollar platform giants — could execute on this vision was the question that would define the next decade.
Thirty Billion Hours
In the spring of 2024, as Roku reported those 30.8 billion streaming hours for the first quarter — roughly 380 hours per active account per year, more than an hour per account per day — the company quietly disclosed a metric that said more about its strategic position than any revenue figure. Roku's platform powered the delivery of content from over 100,000 movies and TV episodes, across thousands of streaming channels, to screens in one out of every three American households with a smart TV.
The content companies needed Roku. They needed its reach, its install base, its home screen. And Roku needed them — without Netflix, Disney+, Max, and the rest, the platform was an empty storefront. The relationship was symbiotic and uneasy, a constant negotiation over revenue shares, ad inventory splits, and home screen placement that played out in conference rooms and contract renewals that rarely made headlines.
Somewhere in San Jose, in an unglamorous office, an engineer's company was collecting a toll on the most valuable thing in the modern media economy — the moment between turning on the television and choosing what to watch. The moment was measured in seconds. The revenue it generated was measured in billions. And the company that owned it was still, after sixteen years, losing money on the box.
