Meta's Engaged-View Attribution Expansion Is Quietly Inflating Your Video ROAS by 15 to 25%
Watch five seconds of a video ad, buy something within a day, and Meta now counts that as an ad-driven conversion.
No click. No swipe. Nothing but attention.
That’s engaged-view attribution folded into Meta’s broader engage-through category, and the expansion earlier this year is already padding reported Meta ads ROAS accuracy by an estimated 15 to 25% on accounts running video creative.
It doesn’t matter if that person closed the app, went about their day, and bought two hours later because a friend mentioned the product in a group chat.
If they watched five seconds of your video before that happened, Meta gives your ad the credit anyway.
What Actually Changed
Meta announced the update on March 3, 2026, under the heading “Simplifying Ad Measurement for a Social-First World.”
The change renamed the old engaged-view attribution bucket to engage-through attribution and expanded what falls inside it.
The video engaged-view threshold dropped from 10 seconds to 5 seconds, and engage-through now also folds in non-link interactions, likes, shares, saves, and comments, that used to count as click-through conversions.
Meta’s own stated reason for the shorter video threshold: the company says 46 percent of Reels purchase conversions now happen within the first two seconds of attention.
The practical effect is that a lot more video views clear the bar for attribution credit than before.
- A 5-second engaged view,
- Or 97 percent of total runtime for anything shorter than 5 seconds,
- Followed by a purchase within a single day,
- Now gets folded into your reported conversions and your reported ROAS, alongside actual click-through sales.
Meta also expanded where engage-through attribution applies, extending it beyond conversion-optimized campaigns into value-optimized Advantage+ Shopping, Catalog, and App campaigns.
Why Short-Form Video Gets Hit Hardest
The math here isn’t subtle.
Most short-form video ads, the 15-second-and-under Reels and Stories format that dominates a lot of e-commerce creative right now, clear a 5-second watch threshold almost by default.
Someone doesn’t have to be persuaded, convinced, or even paying close attention. They just have to not scroll away in the first third of the clip.
That’s a low bar for something Meta is willing to credit as the reason a purchase happened.
Compare that to the old 10-second threshold, which required roughly double the attention span to qualify, and it’s easy to see why the shift lands hardest on exactly the creative format most brands have been scaling all year.
A campaign running short-form video can look like it suddenly got much more efficient, when what actually changed is how generously Meta is willing to hand out credit for a five-second glance.
Creative Win or Measurement Artifact?
Here’s the question worth asking before scaling any video campaign that’s showing a ROAS jump right now: did the creative get better, or did the measurement get looser?
Those are two very different explanations for the same number going up, and only one of them means you should spend more.
A genuine creative win means more people are actually being persuaded by your ad and buying because of it.
A measurement artifact means Meta started counting purchases it was always going to get anyway, the mom-in-the-group-chat kind of sale, and crediting your video for influence it didn’t have. Scaling budget based on the second explanation doesn’t grow revenue. It just grows spend against a number that was never real in the first place.
How to Tell Which One You’re Looking At
Meta’s Ads Manager lets you break out attribution by window, so the fastest gut-check is comparing your account’s reported ROAS under different settings side by side:
- Pull ROAS using only click-through attribution, no engaged-view or engage-through credit included, and compare it against the blended number you’ve been reporting
- Check the gap specifically on video-heavy campaigns versus static image or carousel campaigns, since the inflation concentrates almost entirely in video
- Look at whether the ROAS jump lines up with a creative change you made, or with the March 2026 rollout window, since a jump with no creative change is the clearer signal something structural shifted
- Compare Meta’s reported conversions against your own first-party, click-based conversion count for the same period, since the size of that gap is roughly the size of the inflation
None of this means engaged-view attention is worthless as a signal.
A five-second watch is real engagement, and it can be a genuine leading indicator of interest.
The problem is treating it as equivalent to a click when you’re deciding whether to scale a budget, since Meta’s own reporting doesn’t distinguish between the two once they’re blended into one ROAS number.
Why This Is the Same Problem, Different Mechanism
This is the platforms-grading-their-own-homework problem showing up in a new form.
The same dynamic that inflates ROAS through view-through conversions with no interaction at all is now doing the same thing with engaged-view credit, a slightly higher bar than a pure impression, but still a long way from a click.
Meta has every incentive to count as much of a sale as it can justify, and engage-through attribution is the latest version of that incentive playing out in your reporting. It’s part of the same broader shift covered in how Meta redefined what counts as a click this year.
- Independent, first-party, click-only measurement doesn’t have this problem, because it only credits a conversion to a channel when there’s a verifiable click tied to it on your own domain.
- A five-second video glance that ends in a purchase two hours later, driven by a conversation in a group chat rather than the ad itself, simply doesn’t get counted, because there’s no click behind it to count.
Before you scale any video campaign that’s showing a ROAS jump this year, run the comparison above.
If the lift holds up under click-only measurement, the creative is working and the budget increase is earned. If it doesn’t, you’re scaling against a number Meta’s own attribution settings inflated, not a number your business actually generated.
If you want to see what your video campaigns look like under independent, click-only measurement, side by side with what Meta is currently reporting, book a live AdBeacon demo.
—-
FAQ
What is Meta’s engaged-view attribution?
Engaged-view attribution credits a purchase to a video ad when someone watches at least 5 seconds (or 97 percent of the total runtime for shorter videos) and then converts within one day, even without clicking the ad. It’s now part of Meta’s broader engage-through attribution category.
When did Meta change the engaged-view threshold?
Meta announced the change on March 3, 2026, lowering the video engaged-view threshold from 10 seconds to 5 seconds and renaming the category to engage-through attribution, which also now includes non-link interactions like likes, shares, saves, and comments.
How much does engaged-view attribution inflate reported ROAS?
Estimates put the inflation at roughly 15 to 25 percent on accounts running significant video creative, concentrated most heavily in short-form video under 15 seconds, where a large share of views clear the 5-second threshold by default.
How can I see my real video ROAS without engage-through inflation?
Compare Meta’s reported ROAS using only click-through attribution against your blended number, and compare both against a first-party, click-based measurement layer that only credits conversions tied to an actual click on your own domain.