Viewability and completion rate: what they tell you, and what they don’t

Viewability and completion rate show up in nearly every campaign report, and they often end up treated as proof that a campaign worked. They are useful, but they measure something narrower than most people assume. This post explains what each metric actually captures, how it’s measured, how it can be gamed, and how to use it alongside the outcomes you care about.

What viewability measures

An impression is counted when an ad is served. That doesn’t mean anyone could see it: the ad might load below the fold on a page the user never scrolls, or in a background tab. Viewability tries to close that gap by counting only impressions that had a real opportunity to be seen.

The industry standard definitions come from the Media Rating Council (MRC), the US body that accredits ad measurement, working with the IAB:

• Display: at least 50% of the ad’s pixels in view for at least one continuous second.
• Video: at least 50% of the player’s pixels in view for at least two continuous seconds.

Viewability rate is viewable impressions divided by measured impressions. Note the word “measured”: impressions where the measurement script couldn’t run are excluded from the denominator, which is why measurability deserves its own line in your report.

Many buyers set stricter internal thresholds, such as requiring the full ad in view. That’s a reasonable choice, but be clear which definition a report uses before comparing numbers across partners.

What completion rate measures

Completion rate applies to video. It’s the share of video ad starts that play through to the end. Video ads report progress through standard VAST events (VAST is the IAB Tech Lab template that video players and ad servers use to exchange ads), including start, first quartile, midpoint, third quartile and complete.

Those quartile events are more useful than the single completion number. A campaign where most viewers drop at the first quartile tells a different creative story from one where they drop just before the end.

Two details change how you should read completion rate:

• Player context. A skippable pre-roll, a non-skippable ad, an outstream player in an article and a rewarded video in an app all produce very different completion patterns. Compare like with like.
• Audibility and visibility. A video can “complete” while muted, in a small player, or scrolled out of view. Some buyers track completions that were both audible and visible, which is a much stricter signal.

How these metrics get measured

Measurement differs by environment, and each has blind spots.

• Web: Verification scripts or the ad server check the ad’s position in the browser viewport and whether the tab is active. Measurement can fail inside certain iframes or when scripts are blocked.
• Mobile apps: There is no browser, so measurement relies on the app’s SDK. The IAB Tech Lab Open Measurement SDK (OM SDK) gives verification vendors a common way to measure inside apps, provided the app has integrated it.
• CTV: A TV screen doesn’t scroll, so viewability is less meaningful. The bigger questions are whether the TV is on, whether the stream is real, and whether ads play in a legitimate app. Completion rates are also naturally high because ads are often not skippable.

When partners report different numbers for the same campaign, the cause is usually a difference in definitions, measurement coverage or counting rules, not an error. Agree on the source of truth before the campaign starts.

Where the metrics get gamed

Any metric that affects price becomes a target. Viewability and completion rate can be inflated without adding real value, and it’s worth knowing the common patterns:

• Tiny players. Video played in a very small player can technically meet a pixel threshold while being impossible to watch.
• Stacked or hidden ads. Several ads layered in one spot, or ads placed where a user can’t reach them.
• Forced or sticky placements. Units that follow the user or block content to rack up time in view, at the cost of user experience.
• Autoplay out of focus. Muted video that plays through while the user is doing something else.
• Refresh abuse. Refreshing ad slots rapidly while they stay in view.
• Invalid traffic. Bots can be programmed to scroll, pause and “watch” ads to completion.

The pattern is the same in each case: the metric improves while actual attention does not. A report with near-perfect viewability and completion across the board deserves a closer look, not just celebration.

Pairing metrics with outcomes

Viewability and completion rate are best treated as hygiene checks: they tell you an ad had a chance to work. They don’t tell you whether it did. For that, pair them with measures tied to your objective:

1. Awareness campaigns: brand lift studies, search lift and reach against the intended audience.
2. Consideration campaigns: site visits, time on site after exposure and engagement with the landing experience.
3. Performance campaigns: conversions, cost per acquisition and incremental results from holdout tests.

Then look for alignment. If a placement shows strong viewability but no lift or conversions, the inventory may be technically visible but not really watched. If a placement has moderate viewability but drives results, don’t cut it just to hit a rate target.

Environment matters too. Inventory with transparent supply paths, respectful formats and layered fraud checks tends to produce metrics that reflect real people paying attention. Vortex focuses on high viewability and completion within curated, brand-safe inventory, with real-time insight into where ads run, so the numbers can be checked against the actual placements.

If you want to plan campaigns where quality metrics and transparent placements go together, see how Vortex works with advertisers.