Roku: how a scrappy TV OS became America's dominant streaming platform

Executive overview

TV distribution is shifting from cable to the internet, and the operating system controlling the smart TV interface is the new chokepoint. Roku controls ~40–45% of U.S. streaming hours — more than the next three or four platforms combined — ahead of Google, Amazon, and Samsung.

Roku won through low hardware cost, a simple UI, and an open app ecosystem, then locked in a two-sided network effect: content goes where viewers are, viewers go where content is.

The platform that owns the TV OS becomes the new cable distributor — collecting a toll on every subscription and ad dollar flowing to the screen.

Origins and the Netflix spin-out

• Anthony Wood invented the DVR (Replay TV, late 1990s), lost to TiVo, then founded Roku in 2002 to pursue internet-based TV distribution.
• Netflix hired Wood in 2007 to build a streaming player for the living room; Hastings cancelled it weeks before launch once he saw it would conflict with Netflix's role as a content supplier.
• The project was spun out as Roku; Netflix took a ~25% equity stake, then sold it to avoid conflict of interest.
• Launched in 2008 as the "Netflix Player by Roku"; opened an app store in 2009 to become a full content aggregator.

Why Roku beat Apple, Google, and Amazon

• Apple TV launched first (early 2007) at ~$300 with a restricted app ecosystem — too expensive and too closed.
• Roku launched a year later at ~$100 with a purpose-built OS requiring cheaper chips than Android-based rivals.
• Google Chromecast required phone control with no remote — poor user experience.
• Amazon Fire TV succeeded with low prices but faces two structural headwinds: Google blocks major OEMs from licensing any forked Android (which Fire OS is), and Walmart and Target are reluctant to promote a direct competitor's brand.
• Simple, app-first UI kept adoption high; Roku consistently leads on technical performance (lowest video-start failures, least buffering, per Conviva).

The two-segment business model

• Device segment (~$400M in 2022): sold at breakeven or slight loss — a customer acquisition channel, never a profit centre.
• Platform segment (~$2.8B in 2022, ~60% gross margin): monetises 70M+ active accounts through advertising and subscription revenue share.
• Advertising is ~two-thirds of platform revenue and growing faster than subscriptions.
• First-party ads (Roku sells its own inventory): ~80–85% margin.
• Third-party ads (Roku sells on behalf of content partners): ~50% margin.
• Subscriptions: Roku earns ~$2 per $10 sub, either as a take-rate or as gross revenue with content cost in COGS.
• Key investor metric: active accounts × ARPU — mirrors cable subscriber economics.

OS licensing and the OEM flywheel

• Roku began licensing its OS to TV OEMs in 2014; Google TV started 2010, Amazon Fire OS 2017.
• TCL and Hisense went from near-zero U.S. share in 2017 to 20%+ by 2021 after licensing Roku.
• Vizio (own OS) fell from ~30% to low-teens over the same period.
• Roku provides more than software: designs the main board, certifies factories, sources components, maintains updates, and supports retail placements.
• Google leaves update maintenance to OEMs; Amazon's forked Android is blocked by Google from major OEM licensing.

Competitive durability

• Two-out-of-three Samsung TV buyers (Samsung has ~30% TV market share but only ~10% of streaming hours) add a Roku or Fire stick — the built-in OS still underperforms.
• OS markets historically consolidate to one or two standards (IBM/Microsoft in PCs; Android/iOS in phones); Roku is positioned as one of those winners in TV.
• Anthony Wood controls voting shares, protecting long-term strategy from short-term acquisition pressure.

Cash flow and investment priorities

• SG&A is ~75% people (engineers + sales); minimal software capitalisation.
• Content asset purchases for the Roku Channel (~$80–100M annual use of cash) create the main gap between EBITDA and operating cash flow.
• Management targets EBITDA breakeven — all gross profit reinvested into active account growth.

International expansion

• Canada: entered 2016, now largest smart TV OS by units sold.
• Mexico: push from late 2018, now largest TV OS.
• Active expansion in Brazil, UK, Germany — sequentially targeting large TV markets.
• Linear-to-CTV ad shift is structural; ad dollars will follow eyeballs as they already have in distribution share.

Key risks

• Content consolidation: if Netflix, Disney, and YouTube dominate supply, they gain pricing power over distributors like Roku.
• Ad-tech disintermediation: platforms like The Trade Desk aggregate inventory across all TV OSes, reducing Roku's scale advantage for advertisers.
• Google and Amazon resources: Google already subsidises some OEM deals; both have far greater capital reserves.
• Platform shift: if viewing migrates away from the living-room screen, the TV OS loses its leverage point.
• Mitigant: AVOD streaming hours are fragmenting across many players, not concentrating — favouring OS aggregators.

Lessons for builders and investors

• Subsidise adoption first, monetise later — Roku kept take-rates low on Netflix and YouTube to win the land grab; premature monetisation is how eBay lost China to Alibaba.
• Know what to own vs. license: Roku focused on the OS until scale made direct TV manufacturing worthwhile — entering hardware five years earlier would have alienated OEM partners.
• Platform economics look worse than end-state value during the land-grab phase; EBITDA breakeven masks the compounding asset being built.
• Primary sources: Anthony Wood interviews from 2009 onward, SEC filings, earnings transcripts; Amanda Lotz's We Now Disrupt This Broadcast for media context.