Marketers have spent years optimizing creative for clicks, views and hooks, but new performance data suggests the ads winning the attention economy can be remarkably bad at winning the actual economy.
Marketing has developed an attention problem, and it is not necessarily the one everyone keeps talking about. The industry has spent years worrying about declining attention spans, crowded feeds and the increasingly difficult job of getting consumers to stop scrolling.
That has created an enormous machinery devoted to measuring whether an advertisement successfully interrupts someone, with hook rates, views, clicks and click-through rates becoming convenient proxies for whether creative is working. The problem is that getting someone’s attention and getting someone’s money are two very different achievements.
An analysis of 88,329 sales-focused Meta video ads running during the first half of 2026 offers an uncomfortable demonstration of that distinction. Across $122 million in advertising spend and $212 million in purchase value, the categories generating the strongest attention metrics were frequently nowhere near the top when performance was measured by return on ad spend.
The most striking comparison comes from Health & Beauty and Baby & Toddler advertising, where the relationship between clicks and commercial performance appears almost completely inverted. Health & Beauty generated the highest click-through rate among 14 categories at 2.36%, yet produced a ROAS of just 1.82, while Baby & Toddler produced the lowest CTR at 1.37% and the highest ROAS at 4.99.
Put another way, the category generating the fewest clicks produced roughly 2.7 times the return of the category generating the most. That is a difficult result to reconcile with an advertising culture that has spent years treating engagement as evidence that creative is doing its job.
Attention Is an Input, Not an Outcome
Metrics such as CTR and hook rate are attractive because they are immediate, abundant and easy to understand. Creative A gets more people through the first three seconds than Creative B, so Creative A appears to be better, while Campaign X produces more clicks than Campaign Y and therefore appears to be working.
Except the commercial outcome can tell an entirely different story, and the disconnect is not limited to two unusual categories. Toys & Games generated the highest hook rate in the dataset at 28.79% and the fourth-highest CTR at 1.89%, yet finished near the bottom for ROAS at 1.62.
Electronics demonstrated almost the opposite behavior, recording the second-lowest hook rate at 22.21% while delivering the second-highest ROAS at 3.44. The creative apparently struggled to keep people watching but performed extremely well when judged against the metric ultimately connected to revenue.
These results do not mean attention metrics are useless, because an advertisement still needs to reach and engage someone before it can influence behavior. They do suggest marketers should stop treating attention as though it were inherently evidence of effectiveness.
An ad can be entertaining without being persuasive, clickable without being commercially relevant and memorable without producing meaningful business value. Those distinctions become increasingly important when engagement metrics start determining which creative gets scaled and where the next dollar of media investment goes.
The Best Ad Might Be the One Most People Ignore
There is a deeper creative lesson hiding inside these numbers, and it challenges one of digital advertising’s most persistent assumptions. Marketers have increasingly designed advertising to appeal to as many people as possible because the platforms reward content capable of generating immediate engagement.
Hooks become broader, openings become louder, editing becomes faster and creative increasingly borrows the conventions of entertainment because everything is fighting for the same few seconds of attention. That strategy can produce impressive engagement while simultaneously attracting enormous numbers of people who were never plausible customers.
More selective advertising behaves differently because it does not necessarily need everyone to care. A piece of creative aimed precisely at the person most likely to buy may be completely irrelevant to everyone else, producing weaker aggregate engagement while generating substantially stronger commercial performance among the smaller audience that matters.
In that context, a low hook rate is not automatically evidence that creative has failed. It may occasionally mean the advertisement has successfully filtered out people who were unlikely to become customers in the first place.
That changes the creative question from “How do we make more people watch this?” to “Are the right people watching this?” The difference sounds small, but it represents a fundamentally different philosophy of performance marketing.
The Dashboard Can Be Right While the Marketing Is Wrong
Modern advertising organizations have become extraordinarily good at improving whatever number appears prominently on the dashboard. If creative teams are rewarded for hook rate, they will learn how to increase hook rate, while media teams rewarded for CTR will optimize toward clicks.
Platforms designed to maximize those outcomes will become increasingly efficient at finding people likely to produce them. None of that guarantees those people are going to buy anything, however, which is where the distinction between optimization and actual business performance becomes critical.
Proxy metrics become dangerous when organizations gradually begin optimizing the proxy instead of the outcome it was originally supposed to represent. What began as a convenient indicator of consumer interest eventually becomes the objective itself, even when the relationship between that metric and commercial performance becomes increasingly weak.
The consequences become particularly significant when those metrics influence budget allocation. A marketer looking only at engagement could reasonably conclude that a high-CTR advertisement deserves additional investment, but if that advertisement generates weak returns, scaling it simply turns an impressive dashboard into a more expensive commercial problem.
Stop Benchmarking Everything Against Everything Else
The dataset also exposes another weakness in the industry’s obsession with universal benchmarks, because advertising performance varies dramatically depending on what is being sold. Across the 14 categories analyzed, average ROAS ranged from 1.04 for Software to 4.99 for Baby & Toddler, creating a nearly fivefold difference between the extremes.
That makes a single cross-industry benchmark of limited value when determining whether an individual campaign is actually performing well. Consumer motivation, purchase frequency, consideration cycles, price points and category dynamics all influence how people respond to advertising, meaning a number stripped of that context can easily create the wrong impression.
Expecting creative for software, electronics, beauty products and baby products to behave according to the same engagement standards ignores the economic context surrounding each purchase. Marketers should therefore be asking whether their advertising is outperforming the realities of its category rather than whether it beats an attractive number pulled from an industry-wide average.
Q4 Is Where Bad Measurement Gets Expensive
The timing matters because Q4 turns relatively small measurement mistakes into extremely large ones. Holiday budgets increase, creative volumes accelerate and media teams begin scaling the campaigns that appear to have demonstrated the strongest signals earlier in the year.
If those signals are based primarily on clicks, views or hooks, brands risk pouring their largest budgets into advertising that is extremely efficient at generating activity but mediocre at generating revenue. A metric that merely misleads during a relatively quiet month can become extraordinarily expensive once Black Friday and holiday spending push budgets to their annual peaks.
The smarter approach is not to abandon attention metrics but to put them back in their proper place. Hook rate can help diagnose whether creative earns an opportunity to communicate, CTR can help identify whether an advertisement creates enough interest to generate further action, and views can reveal whether media and creative are successfully reaching audiences.
ROAS, conversion, incremental revenue, customer acquisition cost and ultimately profitability answer a different and considerably more important question. They help determine whether any of that attention was actually worth buying, which should ultimately matter more than whether someone stopped scrolling for three seconds.
Stop Optimizing for Applause
This distinction will become even more important as advertising platforms become better at optimizing toward whatever objectives marketers give them. Machines can become extraordinarily efficient at generating clicks, views and engagement, but that efficiency becomes dangerous when the objective itself is poorly chosen.
If marketers tell an increasingly sophisticated advertising system to find people who click, it will find people who click. That does not necessarily mean it will find people who buy, become profitable customers or generate enough lifetime value to justify what it cost to acquire their attention.
The next era of performance marketing may therefore require marketers to become less impressed by attention rather than more obsessed with capturing it. The advertisement everyone stops to watch is not necessarily the advertisement that works, while the quieter creative ignored by most of the feed may be doing something far more valuable.
Marketing does not ultimately get paid for stopping thumbs, generating clicks or producing impressive-looking dashboards. It gets paid for changing behavior and creating economic value, and the difference between those two things may be considerably larger than the industry’s favorite metrics would have us believe.