The Model No One Saw Coming
When Roku was selling streaming players for $29, the market perceived it as a low-cost hardware manufacturer competing against Amazon and Google. This reading was always misguided. Roku was never a hardware company; it was a company focused on consumer attention, using hardware as bait. The results from Q4 of 2025 confirm this with a mathematical precision that deserves more analysis than it has received.
The company reported $1.39 billion in quarterly revenue, exceeding estimates by 3%, with platform revenue growing by 25% year-over-year and crossing the $1 billion threshold in a single quarter for the first time. GAAP EPS came in at $0.53, 88% above consensus. These are not the numbers of a company selling screens. They are the numbers of a company selling audiences.
What CEO Anthony Wood described as a "powerful tailwind" from artificial intelligence is not corporate fluff; it is a technical description of how Roku is compressing the time between a user turning on their TV and a brand paying Roku for that moment of attention. Each content discovery improvement, every personalized recommendation, every programmatic ad format activated through platforms like Amazon, AppLovin, or The Trade Desk eliminates friction on the advertiser's side. And friction removed from the buyer's side always becomes sustainable pricing.
The Arithmetic Behind International Expansion
The number structuring Roku's entire thesis toward 2028 is not total revenue. It is ARPU —average revenue per user. And here is the mechanics that few are articulating clearly enough.
In mature markets like the United States, where Roku already exceeds half of broadband households, ARPU has scaled from $28.76 to over $41 in recent years. In Mexico, Brazil, Canada, and the United Kingdom, that figure is materially lower. Not because consumers value entertainment less, but because local advertising markets still lack the density of advertisers, programmatic infrastructure, and inventory maturity found in the U.S. market.
That is precisely the opportunity. Roku has already solved the hardest problem: mass installation. The Roku TV operating system was the best-selling in Canada and Mexico during Q4 of 2024. Partnerships with Sanyo in Mexico, Philco, TCL, and SEMP in Brazil, and AOC in Chile and Peru have built a base of connected homes. Retail distribution anchored in Best Buy, Target, Amazon, and Walmart, with expansion to Walmart physical stores since February 2025. The company possesses the asset. What comes next is the monetization of that asset, which historically generates disproportionate returns because the marginal costs of activating an advertiser on an already installed screen are nearing zero.
CFO Dan Jedda articulated it with surgical precision: double-digit growth in platform revenue sustained by disciplined operating expenditure growth and stable gross margins. That combination—revenue expansion with cost containment—is the operational definition of leverage. It is not an optimistic projection; it is the mechanical consequence of having a built distribution network and starting to charge for access to it.
Why the Premium Subscription Model Matters More Than It Appears
There is a component of Roku's expansion that has received insufficient coverage: the growth of premium subscriptions via The Roku Channel, which recorded a net record of new subscribers in Q4, including the launch of services like Frndly and Howdy. This is not a secondary detail. It is the piece that transforms Roku from an advertising network into a multi-revenue platform.
When Roku sells a premium subscription through its own channel, it captures a percentage of that transaction with a high margin and none of the customer acquisition costs that an independent service would endure. The user is already on the platform. The screen is already on. The algorithm already knows what interests them. The friction of converting a casual user to a paying subscriber is structurally lower than in any other channel. That is what differentiates a platform from a product: the platform reduces the cost of each additional transaction because the context is already built.
The expansion of Howdy to non-Roku platforms is particularly relevant from a competitive positioning perspective. If Roku can distribute content and capture subscription revenue outside of its own hardware, it begins to operate more like a media channel than a device manufacturer. That move, if executed, alters the business's risk profile because it decouples revenue from television sales.
The Certainty Gap Roku Still Needs to Address
The risks are not hidden; they lie in the assumptions behind the guidance of $1.2 billion for Q1 2026 and the projection of over $1 billion in free cash flow for 2028. Each of those numbers depends on international monetization scaling at a pace compatible with the installed user base.
The connected TV advertising market in the United States projects between $40 and $50 billion by 2027. The international markets where Roku is growing are structurally less mature: fewer sophisticated digital advertisers, less programmatic budgets, volatile currencies, and data regulations varying by jurisdiction. Roku may have the audience inventory. But if local advertisers lack the tools or budget to buy it, the international ARPU will not converge with North American levels as quickly as the company's models suggest.
Moreover, Amazon Fire TV, Google TV, Samsung Tizen, and LG webOS are not passive competitors. All have their own incentives to build closed advertising ecosystems that capture the same advertiser budgets. Roku got there first and has scale in North America. In international markets, that first-mover advantage is thinner.
Zacks' estimate of an EPS of $2.10 for 2026—which would imply a 255% growth—is not an implausible forecast if platform expansion continues. But it is a projection that leaves little room for imperfect execution in markets where Roku is still building advertising demand infrastructure.
ARPU Doesn't Rise Just Because You Have More Users
This is the structural lesson that Roku's expansion offers to any business operating under a platform model: the installed user base does not automatically monetize. What monetizes is the perceived certainty of the outcome by the advertiser or subscriber deciding to pay. Roku in Mexico has screens. But the Mexican advertiser needs to believe, with data, that screen delivers the customer they seek, with the traceability they need, at a price that justifies the budget.
Integration with demand platforms like The Trade Desk and AppLovin is precisely that mechanism: it reduces the advertiser's friction in buying inventory and increases their certainty about the outcome. Each such connection is not just an additional revenue source; it is a reduction in the effort it costs the buyer to access the inventory, which sustains higher prices and more stable margins over time.
The trajectory towards 100 million homes and $1 billion in free cash flow is not a bet on distribution. It is a bet on whether Roku can replicate, in international markets with lower advertising density, the same architecture of certainty it built in the United States. The Q4 numbers of 2025 indicate that the company knows how to do it. What remains to be proven is if it can do so at the pace its own projections demand.
Sustainable commercial success, in platforms as in any other model, always results from the same equation: minimize buyer friction, maximize their certainty about the outcome they will obtain, and build a proposition whose perceived value makes the alternative of not paying seem irrational. That is what Roku is executing. The question is not whether the model works; it's whether the speed of execution in international markets reaches before the competition closes the gap.
