How much is one more scroll worth? Meta trial puts attention economics under scrutiny

Being an adult has its ups and downs, but almost always comes with optionality. The next mobile game, the next poker hand, the next dating app. We all participate, and then we (theoretically) have the wherewithal to stop.

Children can’t choose nearly so well.

That distinction now sits somewhere near the heart of one of the most consequential trials Meta has faced. In an Oakland federal court, California, Colorado, Kentucky and New Jersey, among 29 US states in total, are arguing that Facebook and Instagram did not merely host material that could harm young users, but were designed in ways that prolonged their engagement despite Meta allegedly knowing more about their particular vulnerabilities than it publicly acknowledged.

What makes the case particularly interesting for the advertising industry is that it is not principally about one offensive post, one dangerous creator or even one badly targeted advertisement.

It is about the architecture around them.

Infinite scroll. Likes. Recommendation systems. Time limits. The mechanics that make the next piece of content easier to consume than stopping. The states want the court to order changes that could include removing infinite scroll and likes for younger users, imposing time limits and strengthening mechanisms intended to keep under-13s off the platforms. Attorneys general indicated before the trial that potential damages could be around $200 billion, although Meta has argued its theoretical exposure could run much higher.

There is nothing inherently sinister about reducing friction. Indeed, much of technology’s history is a history of making the next action easier. Search got faster. Payments became one-click. Television stopped asking viewers to wait a week. Social media simply became extremely good at ensuring that the next choice is already sitting there.

For adults, that is largely a trade we make ourselves. The court is now being asked whether the same bargain looks different when the user is a child.

When engagement itself becomes the question

California Deputy Attorney General Megan O’Neill opened the states’ case by arguing that Meta had extensively studied how young people respond to rewards, social feedback and impulse, and then built products around that knowledge. Former Meta safety engineer Arturo Béjar, the states’ first witness, testified that safety considerations were sometimes diminished even when engineers had ideas for improving products. Meta attorney Paul Schmidt countered that the company had invested extensively in safety and argued that the evidence would not establish the causal relationship between adolescent social media use and poor well-being that the states allege.

That legal argument has an advertising corollary.

A feed produces inventory. More time produces more opportunities to show ads. Recommendation systems find the piece of content most likely to keep the session alive. The next swipe requires almost no deliberation at all.

The advertising business did not invent that behaviour, nor does every extra minute on Instagram automatically represent unhealthy engagement. But if a court begins drawing a legal distinction between ordinary product functionality and engagement-maximising design when children are involved, one of digital advertising’s oldest assumptions becomes less comfortable: that more attention is, by itself, an uncomplicated good.

Read more on attention economics here 

Veteran marketer Shubhranshu Singh believes the immediate revenue effect could actually be smaller than the headline numbers imply. Teen impressions are already relatively lightly monetised, he argues, with Meta having restricted behavioural and interest-based targeting for younger users and several high-value advertising categories unable to target minors.

The larger question is what happens later.

“Teens age into higher ARPU adults, with engagement habits set early,” Singh notes. If product changes blunt that habit formation, he argues, Meta potentially loses something more valuable than a few low-yield teenage impressions: the pipeline into highly engaged adult users.

That is particularly relevant to Reels. Autoplay and an effectively endless feed are not incidental to short-form video; they are central to how the format is consumed. Singh therefore sees any restriction on those mechanics as potentially biting hardest into the product Meta built to answer TikTok.

There is a second-order problem too. Digital platforms have several ways to offset declining inventory, including increasing ad load. But if the problem being adjudicated is excessive engagement itself, simply squeezing more advertising into the remaining attention becomes a less elegant answer.

Fewer minutes, better minutes?

Gopa Menon, Co-founder and COO of theblurr, expects the simpler arithmetic to show up first.

“If Meta is forced to reduce engagement-maximising features for younger users, available ad inventory in those cohorts is likely to fall,” he says, potentially putting some pressure on campaign volumes and pricing.

But Menon sees another possible consequence: advertisers being forced to distinguish between attention and useful attention.

If raw time-spent metrics become less abundant, he argues, brands may place greater emphasis on outcomes, meaningful engagement and quality of reach rather than treating every additional minute inside a platform as equally valuable.

That distinction matters because the states’ argument is not that social media has no value. Even O’Neill acknowledged during opening statements that it can have benefits. The question is whether there is a point at which the mechanisms used to maximise continued participation become a product-safety issue for users who have less developed capacity to disengage.

Meta, naturally, contests that characterisation. Its defence says internal conversations are being presented without sufficient context and that attempts by employees to discuss and improve products should not themselves be treated as evidence that those products were knowingly dangerous. The company has also repeatedly pointed to protections it has introduced for teenagers.

That distinction will be for the court, not advertisers, to resolve. But advertisers may eventually have to deal with the consequences.

And then there is India

India makes this more than a distant American courtroom story.

It is Meta’s biggest market by user numbers. Meta does not disclose country-level user totals, but DataReportal estimates cited by Reuters earlier this year put Facebook’s Indian audience at about 403 million and Instagram’s at roughly 481 million.

Read more on Meta in India 

India is also already approaching youth protection from a different regulatory direction.

Section 9 of the Digital Personal Data Protection Act bars processing likely to have a detrimental effect on a child’s well-being and prohibits tracking or behavioural monitoring of children and targeted advertising directed at them. Under the law, a child is a person below 18. The DPDP Rules notified in November 2025 establish mechanisms for verifiable parental consent, with the relevant provisions scheduled to phase in after the prescribed implementation period.

Singh puts the difference neatly: “The Indian lever bites monetisation, the US lever bites the mechanics of habit.”

A ruling in Oakland would, of course, not become Indian law.

Ekta Rai, Advocate at the Delhi High Court, says as much. But she argues that the evidence emerging from the case could still matter. If Meta’s own internal material establishes that specific engagement-maximising features were retained despite known risks to children, that evidence could potentially become relevant to Indian regulators or courts considering the country’s existing child-data and consumer-protection framework.

Her question is less about importing American jurisprudence than explaining different treatment: if Meta is eventually required to build stronger safeguards for children in the US, why should the same underlying product risks justify materially weaker protections for children in India?

Anshul Verma, Partner, SKV Law Offices, makes a similar distinction around platform liability. The lawyers argue that the US case is partly testing whether claims aimed at product design — including feed architecture, prompts and notification systems — can be separated from claims concerning third-party content. India’s conditional intermediary safe-harbour regime is structured differently, they note, while the DPDP framework already provides separate tools around children’s data and behavioural monitoring.

That does not mean Indian brands suddenly inherit legal liability because somebody in California dislikes infinite scroll.

It does mean the governance question could become harder to ignore.

Menon expects global safety standards to travel even when regulation does not travel at the same speed. If a major jurisdiction ultimately determines that particular engagement mechanics require stronger protections for minors, brands operating across markets may start asking not simply how much engagement they are buying, but what produced it.

For an advertising ecosystem accustomed to treating attention as scarce inventory, that is a fairly fundamental question.

What makes an ad impression worth in an AI conversation? Read here

The Meta trial may ultimately vindicate the company, punish it financially, force product changes, or arrive somewhere between those outcomes. It is expected to run for around six weeks, with Mark Zuckerberg and Instagram chief Adam Mosseri among those expected to testify.

But the commercial question raised on Day One is already larger than Meta.

Digital products have spent two decades becoming exceptionally good at making the next action easy. That is usually called good design. For adults, the next scroll, hand, match or video remains a choice, however frictionless that choice may become.

The trial in Oakland is asking what happens when the person making that choice is a child.

And if the answer changes the design of the feed, the economics built on top of that feed may have to change with it.