Meta reduced the engaged-view attribution threshold from 10 seconds to 5 seconds for video ads. Conversion counts rise and CPA drops on paper, but actual campaign performance has not changed. Advertisers need to recalibrate benchmarks.
If your ROAS jumped recently and you cannot explain why, check your attribution settings before you celebrate.
Meta has cut the engaged-view attribution window for video ads from 10 seconds to 5 seconds. The change means more viewers now qualify as "engaged" before dropping off, which inflates conversion counts and deflates cost-per-acquisition figures in reporting.
Nothing about actual campaign performance has changed. The same viewers are watching. The same conversions are happening. Meta has simply lowered the bar for what counts as an engaged view.
Meta cited behavioural data to justify the change. In a Reels-dominant environment, 46% of online purchase conversions happen within the first two seconds of viewer attention on a video ad. The new five-second threshold is intended to be a more accurate indicator of engagement, according to Meta.
The broader attribution overhaul also renamed "engaged-view" to "engage-through" attribution, aligning it with changes Meta made to click-through attribution in March.
Of online purchase conversions with Reels ads happen within the first two seconds of viewer attention, per Meta
Why it matters
For Australian advertisers running video campaigns on Meta, this change has immediate reporting implications. If you run video ads between 6 and 15 seconds long, many viewers who previously dropped off between 5 and 10 seconds now count as "engaged." Your conversion count rises. Your CPA drops. Your ROAS improves. None of this reflects a real change in performance.
The risk is that teams make budget decisions based on inflated metrics. A campaign that looks 15% more efficient today than it did last month may not have improved at all. The measurement moved, not the outcome.
Meta is clear that billing is unaffected. You are not paying more or less. This is purely a reporting change. But reporting drives decisions, and decisions drive spend. Inflated metrics lead to overconfidence, which leads to over-investment in channels that may not be delivering what the numbers suggest.
What to do about it
Meta is not wrong that short-form video engagement happens fast. But cutting the measurement window inflates metrics without improving outcomes. The advertisers who win are the ones who know the difference.