Viewability measures whether an ad was actually seen, not just served. The common industry standard counts an impression as viewable when at least 50 percent of the ad's pixels are on screen for at least one second, though standards vary by format and measurement vendor. Low viewability means a publisher or advertiser is paying for ads that no one had a real chance to see.

How to calculate viewability

Viewability rate = (Viewable impressions / Total measured impressions) x 100

Viewability rate = (Viewable impressions / Total measured impressions) x 100

Viewability rate = (Viewable impressions / Total measured impressions) x 100

The rate is the share of measured impressions that met the viewable threshold. A measured impression is one where a viewability tool could actually check the ad's position on screen and its time in view, which is why "measured" matters: impressions that cannot be measured are excluded from the denominator. Formats and vendors set their own thresholds, so a video standard may require a longer time in view than a static display unit, and two tools can report slightly different rates for the same placement.

A worked example

Say a placement records 10,000 measured impressions in a day, and 8,000 of them met the 50-percent-for-one-second threshold:

(8,000 / 10,000) x 100 = 80% viewability

(8,000 / 10,000) x 100 = 80% viewability

(8,000 / 10,000) x 100 = 80% viewability

An 80 percent viewability rate means four out of five served-and-measured ads had a genuine chance to be seen. The remaining 20 percent scrolled past too quickly, loaded below the fold, or were closed before the threshold was met.

Why viewability matters

An Ad Impression that is never seen has no value to the advertiser and, over time, no value to the publisher either. Advertisers increasingly buy on viewable impressions and judge inventory by its viewability rate, so a placement with poor viewability attracts lower bids. It also distorts other metrics: CPM and eCPM look artificially efficient when they are calculated over impressions that were never actually in view. Measuring viewability keeps the whole picture honest.

What hurts viewability

• Below-the-fold placement. Ads that load where users rarely scroll may never enter view.

• Fast scrolling. If a user passes the ad before the one-second threshold, it does not count as viewable.

• Slow loading. An ad that renders late can miss its chance to be seen at all.

Why offerwalls sidestep the viewability problem

Viewability is mostly a problem for passive display formats that are served and hoped-for. An offerwall works differently, because the user opens it deliberately and actively engages with the offers on it. There is no question of whether the content was on screen long enough: the user is looking at the wall, choosing an offer, and completing it. Offerwall earnings are tied to completed actions rather than served impressions, so the gap between "served" and "seen" that viewability exists to catch barely applies. That also makes offerwall value easier to trust than a raw Reach or impression count.

Common mistakes to avoid

• Treating served impressions as seen. Without a viewability measure, impression counts overstate how many ads reached anyone.

• Chasing 100 percent viewability. Some unmeasured or missed impressions are normal, and forcing the number can hurt user experience.

• Reading eCPM without viewability. A high eCPM on low-viewability inventory can be worth less than it looks.