Meta expects to record an approximately $10 billion legal expense in the third quarter after agreeing to restrictions on how Facebook and Instagram capture teen attention. During the same week, YouTube fought a UK proposal that could interfere with its recommendation system, the BBC proposed building a “public service algorithm,” Flipboard bought technology that lets outside curators control and monetize their own feeds, and Roku converted sports discovery into exclusive NWSL rights.
Recommendation now determines enough economic value that companies, regulators, creators, rights owners and public institutions are fighting over who gets to control it.
The companies controlling what gets recommended can allocate audience, reduce customer acquisition costs, create advertising inventory, direct transactions and decide which suppliers receive meaningful distribution. That control also carries costs. Product restrictions can reduce available attention. Prominence rules can reallocate impressions. Curators can demand a share of advertising economics. Rights owners can expect distribution commitments alongside rights fees.
Meta Put a Price on the Engagement Loop
Meta’s proposed settlement with a bipartisan coalition of attorneys general turns engagement mechanics into enforceable operating restrictions.
Teen users will face default daily limits across Facebook and Instagram, restricted nighttime access, muted notifications during school hours, stronger age-assurance requirements and additional parental controls. Most provisions remain in effect for ten years. Meta expects the agreement to produce an approximately $10 billion legal expense in Q3.
Push notifications generate return sessions. Personalized recommendations reduce the friction between one piece of content and the next. Continuous feeds extend sessions. More time creates more opportunities to serve advertising.
Meta’s settlement places legal limits on those engagement mechanics.
The revenue effect won’t track directly with the amount of viewing restricted. Teen users represent only part of Meta’s business, parents can change some defaults, and several activities fall outside the limits. The operating constraint is still new. Meta can no longer optimize every eligible teen session around maximizing additional time.
Higher advertising yield, stronger commerce conversion, better creator monetization and formats capable of generating value without extending feed time become more important when attention itself has a ceiling.
The settlement also gives Meta a financial incentive to spread the restrictions across the category. Roughly $5.3 billion of the company’s stated payment obligation depends on YouTube and TikTok implementing specified limits and making corresponding payments. If its largest competitors operate under similar rules, Meta reduces the competitive penalty attached to limiting its own users.
YouTube Is Fighting Over Who Gets to Allocate Demand
The UK government is considering ways to give public-service media and trusted news greater prominence inside digital products including YouTube. Its preference is for voluntary agreements, with legislation remaining an option if those agreements don’t produce enough visibility.
Pedro Pina, YouTube’s vice president for EMEA, used the Edinburgh TV Festival to argue that mandated prominence would displace creators and publishers who currently compete for recommendations through YouTube’s existing system.
YouTube’s fight over UK prominence rules is fundamentally a fight over recommendation control.
There may be effectively unlimited video supply on YouTube, but the home screen, recommendation rail and next-video slot are finite distribution inventory. Giving one supplier additional prominence reduces the inventory available to another.
A recommendation can produce views, advertising revenue, subscriptions, channel growth and downstream transactions. A creator with millions of uploaded videos competing against it doesn’t receive meaningful economic value from theoretical availability. The business begins when the recommendation system puts the video in front of somebody likely to watch.
Traditional carriage negotiations were explicit about placement. Networks fought for channel positions, tier inclusion and marketing commitments because distribution affected audience and affiliate economics.
Algorithmic distribution made those decisions personalized and continuous. It didn’t remove the underlying economic fight.
The BBC Wants to Compete With the Incentive Behind the Feed
BBC Director-General Matt Brittin has proposed opening iPlayer to more creator-led programming and using a “public service algorithm” to recommend it.
The concept would give the BBC a recommendation system designed around its public mission rather than maximizing advertising or subscription economics. It would also turn the BBC into something more complicated than a commissioner and distributor. The corporation would operate the environment, establish participation rules and decide which outside programming receives exposure.
The BBC proposal would put creator supply and algorithmic recommendation inside iPlayer.
Brittin says the BBC has lost approximately £1.3 billion in revenue during the past decade. The corporation is pursuing £500 million in annual savings and as many as 2,000 job reductions. Creator supply offers a way to expand the range of programming available inside iPlayer without financing every hour through a conventional commissioning model.
An open creator system requires moderation, rights management, ranking policies, payment structures, appeals and rules governing distribution. Public-service standards add obligations that a purely commercial recommendation system doesn’t carry in the same form.
YouTube’s ranking system creates value because advertisers, creators and viewers participate in the same commercial loop. The BBC wants personalization without maximizing commercial engagement while still paying for the infrastructure, moderation and programming supply required to make the product competitive.
Flipboard Is Splitting the Economics of the Feed
Graze lets creators, publishers and communities build custom social feeds, select their sources, establish ranking rules and sell contextual advertising against the resulting audience. Graze says its technology delivered more than 41 billion posts to roughly 12 million people during its first 21 months, with more than 7,000 feeds and approximately 60% of traffic monetized.
Advertising revenue is split 70/30, with the majority going to the feed creator.
Flipboard is betting that the person programming a feed can own more of its economics.
A large social platform supplies the interface, recommendation system, audience data, ad marketplace and payment structure. The creator supplies programming but has limited control over how the surrounding feed behaves or how its economics are divided.
Graze gives a curator control over more of the programming decision and a direct share of the advertising revenue attached to that decision.
If ranking can be separated from the company hosting the underlying content, recommendation becomes its own product.
A women’s sports publisher could build a feed around selected journalists, teams, creators and video sources. A league could organize the conversation around its own programming. A media company could build a specialized feed without depending entirely on the default ranking system of the social service where those posts originate.
The party assembling demand can then ask to participate in the economics produced by that demand.
Roku Shows Why Recommendation Control Is Worth Owning
Roku launched an NWSL Zone in 2025 to organize matches, schedules, highlights and viewing destinations across the league’s fragmented distribution. It will now carry eight exclusive Sunday Night matches during the remainder of 2026 and 25 in 2027 on the free Roku Sports Channel.
Roku turned an NWSL discovery relationship into exclusive rights and advertising inventory.
Roku can promote the match on its home screen, route the viewer into The Roku Channel, sell the resulting advertising inventory and measure what the household does before and after the game.
Roku’s position at the point of discovery gives it something a conventional rights buyer has less of: direct behavioral evidence about the audience it is considering acquiring.
Roku says streaming hours inside its North American Sports Zones increased 122% year over year during Q2 while unique visitors rose 80%. Those company figures don’t disclose the revenue generated by the zones, but they establish that more sports consumption is beginning inside the interface Roku controls.
Exclusive rights let Roku monetize more of the demand its recommendation environment creates.
YouTube is applying a related model to programming it already hosts. Its expanded Stations experiment packages creator videos, media channels and podcasts into continuous streams surfaced through Home and Search recommendations. The service can create additional lean-back inventory from existing supply without commissioning a conventional network schedule.
YouTube can turn already-uploaded catalogs into programmed channels.
When the distributor recommending programming also owns that programming, rights owners, regulators and competing suppliers have reason to scrutinize how the interface allocates exposure.
The Streaming Wars Take
Prominence is becoming a commercial term.
Sports leagues selling rights will have reason to negotiate where matches appear inside home screens and sports hubs, how often they’re promoted, which audience data comes back and whether the distributor can favor inventory it owns. Streaming services entering aggregation deals will care about recommendation treatment as much as app availability. Creators participating in feed businesses will push for visibility rules and revenue participation. Regulators will continue defining where commercial optimization ends for protected audiences and public-interest programming.
A contractual promise to “promote” a service has limited value without agreed definitions for impressions, placement, targeting, frequency and audience delivery. The more personalized the interface becomes, the harder it is to verify whether two suppliers received comparable distribution.
Recommendation data therefore becomes part of deal economics. Rights fees, revenue shares and carriage terms can be evaluated against the audience the distributor can actually produce rather than the theoretical reach of its installed base.
The companies controlling the interface gain bargaining power because they can steer demand toward inventory at lower acquisition cost. Suppliers gain leverage when their programming is valuable enough to make that interface less useful without them. Regulators can alter both sides by reserving attention for particular audiences or suppliers.
Recommendation now allocates revenue, determines supplier economics, shapes rights value and creates regulatory exposure. That puts the algorithm in the P&L, the contract and the negotiation.
The Streaming Wars is intentionally ad-free
We don’t run display ads. Not because we can’t, but because we don’t believe in them.
They interrupt the reading experience. They cheapen the work. And they burn advertisers’ money on impressions nobody actually wants.
So we chose a different model.
We say the things people in this industry are already thinking but don’t say out loud. We connect the dots beyond the headline and focus on explaining why things matter to the people working in this business.
If you believe industry coverage can exist without clutter and interruption, you can support it here → SUPPORT TSW.
Support is optional. But it directly funds research and continued coverage — and helps prove this model can work.
Support TSW →
