View rate calculator

Video views is the platform's own count, on whatever definition that platform uses: two seconds, three seconds, six, ten, a share of the video's length, or a completion. Some placements play automatically, which makes a view nearly the same event as an impression; others require a deliberate tap. Take the definition from the report before you take the number.

Impressions is the number of times the video was served.

Ad spend is optional and produces the two price lines, which is where most of the value on this page is.

A rate is only comparable against the same definition on the same surface.

An autoplaying, full-screen placement will produce very high view rates for every advertiser on it, because the video starts playing whether anybody chose it or not. A feed where video starts muted in a small frame will produce low ones for everybody, because the first second has to earn the second. Two campaigns reporting eighty and fifteen may be the same creative bought in two places.

The useful comparison is internal: your creative against your creative, on one placement, over a stable period. That comparison answers a real question — did the new opening work better than the old one — and it is the only version of this metric that can.

The relationship is exact and it settles most arguments about video performance. The cost of a view is the cost of the impression divided by the rate at which impressions became views.

That means a rising cost per view has exactly two possible causes, and the page tells you which. If the cost per thousand impressions rose and the rate held, the market got more expensive and nothing about the creative changed. If the rate fell and the price of the impression held, the creative stopped stopping people. The first is a budget and timing conversation; the second is an editing conversation, and treating one as the other wastes a month.

The same relationship explains why buying the cheapest inventory rarely produces cheaper views. Low-attention placements are cheap per thousand and convert far fewer impressions into views, so the price per view lands in the same place or worse, with more impressions burnt to get there.

The first second, more than everything else combined. On a muted, scrolling surface the opening frame is the whole decision. Most view-rate problems are opening-frame problems, and the cheapest test available is three openings on one edit.

Sound design. A video that needs sound to make sense will lose most of an audience that never turns it on. Captions and a visual that carries the meaning without audio are not accessibility extras on a feed; they are the format.

Aspect ratio and framing. A video cropped from a wide edit into a vertical placement wastes most of the screen, and the wasted screen is the part that was meant to stop somebody.

Length, but less than people think. A long video with a strong opening holds better than a short one with a weak one. Length matters far more to the hold rate than to this number.

The audience. A rate that is poor across every creative is usually a targeting result rather than a production one.

Autoplay inflates it. On surfaces where the video plays by itself, a view is close to an impression and the rate says almost nothing about intent.

It ignores everything after the view. A rate of thirty per cent with nobody watching past the first few seconds is a hook that worked and a video that did not. Read the hold rate beside it.

It is not a funnel step. A view is not an interest signal in the way a click is; it is frequently an accident of scrolling speed. Campaigns that treat accumulated views as an audience to retarget are often retargeting people who never chose to watch anything.

It cannot be added across platforms. Different definitions, different surfaces, different autoplay behaviour. A blended view rate across an account is a number with no referent.

Use the rate to compare your own creatives on one placement, use the cost per view to compare periods, and use the cost per thousand beside it to work out which of the two moved. Set a target from your own trailing figures rather than from a published average, because the published average is a blend of surfaces that do not resemble yours.

And keep the objective in view. A view is a cheap outcome to buy and rarely the one the business wants. Where the campaign exists to sell something, this number is a diagnostic for the creative and the decision still belongs with the cost of the sale.

One habit is worth adopting whatever the numbers say: record the definition beside the figure. A view rate written down without the platform, the placement and the seconds behind it is a number that cannot be checked six months later, and most of the disagreements about video performance are two people quoting different definitions at each other in good faith.