How an engaged view is counted
Google counts an engaged view when someone watches at least ten seconds of a video ad, or watches the whole advert when it runs shorter than ten seconds. It sits deliberately between two other thresholds: an impression, which only records that the advert was served, and a full view on skippable in-stream, which asks for considerably more watching or an interaction.
The metric earns its keep because so much video is now short and served in feeds. On formats where almost nobody watches to a long threshold, a report built only on views would show very little happening. The engaged view is the platform’s answer: a lower bar that still excludes the people who scrolled straight past.
Why engaged views matter
They are the counting unit behind engaged-view conversions, which credit a sale or enquiry to someone who watched past that threshold, did not click, and converted within a set window afterwards. That is how video gets any credit at all in campaigns where the click is not the natural next step — and it is also where careful reading becomes essential.
The count is also a practical guide to creative length. If the gap between impressions and engaged views is wide, the opening is not holding anyone, and no amount of budget will fix a video that loses people before the threshold is reached.
Where engaged views go wrong
The first problem is quiet inclusion. Engaged-view conversions can be folded into a headline conversion total, so a campaign appears to have produced enquiries when a portion of them are credited to people who only watched. That is not fraud and it is not necessarily wrong, but it must be visible in the report, or budget decisions are made on a number the owner does not understand.
The second is reading engagement as intent. Watching a few seconds of an advert in a feed is a very low commitment, and much of it happens with the sound off while somebody decides whether to scroll. It is a signal, not an interest.
The third is comparing engaged views with views as though they measured the same thing. They use different thresholds by design, so any table that ranks campaigns using both columns together is comparing an easier test with a harder one.
How to act on it
Split the reporting first. Keep click-driven conversions and engaged-view conversions in separate, clearly named rows, and look at what remains when the engaged-view row is set aside. If the campaign still stands up, it is genuinely working. If nearly all of the result sits in that row, the honest next step is a test — pausing the campaign in one region, or running an incrementality test — rather than a bigger budget.
Then use the count to improve the creative. Move the brand, the offer and the reason to care ahead of the threshold, keep the message legible without sound, and test lengths against each other rather than against a rule of thumb. Those are the everyday decisions that make Demand Gen campaigns and short-form video worth their budget.